CN110809429A

CN110809429A - System and method for filtering noise and analyzing venous waveform signals

Info

Publication number: CN110809429A
Application number: CN201880043655.3A
Authority: CN
Inventors: 乔纳森·汉德勒; 詹姆士·马尔图奇; 凯尔·霍金; 苏珊·伊格尔; 克里恩·布罗菲; 理查德·博耶尔; 弗兰兹·博登巴赫尔
Original assignee: Baxter Healthcare SA; Baxter International Inc
Current assignee: Baxter Healthcare SA; Baxter International Inc
Priority date: 2017-06-30
Filing date: 2018-06-29
Publication date: 2020-02-18
Also published as: SG11201913638PA; KR20200024855A; JP2020526260A; BR112019027925A2; CO2020000673A2; MX2019015734A; EP3644837A1; WO2019006362A1; US20190000326A1; AU2018294354A1

Abstract

Devices, systems, and methods for filtering medical device noise artifacts from venous waveform signals are disclosed. Peripheral Venous Pressure (PVP) is measured and transformed from the time domain to the frequency domain for analysis to determine patient status. To avoid pumping artifacts, the time domain PVP measurements are filtered to generate a filtered time domain PVP signal by removing active pumping periods. The filtered time-domain PVP signal is transformed into a frequency-domain PVP signal, which is analyzed based on peaks indicative of a respiration rate, a heart rate, or harmonics thereof. A measure of the patient state is then determined from the peak or corresponding frequency. The patient state may be related to the patient's blood volume and may be used to control pump operation.

Description

System and method for filtering noise and analyzing venous waveform signals

Priority requirement

The present application claims priority from: U.S. provisional application No.62/671,108 entitled "System and method for Monitoring and Determining temporary parameter from Sensed vessel wave", filed on 14.5.2018, U.S. provisional application No.62/599,421 entitled "Systems and methods for Filter Medical Device non electrodes from vessel wave for", filed on 15.12.2017, U.S. provisional application No.62/527,944 entitled "System and method for Filter Medical Device non electrodes from vessel wave for", filed on 30.6.2017, and U.S. provisional application No.62/528,570 entitled "System and method for testing Medical Device non electrodes from vessel wave for measuring Signal", filed on 5.7.7, are incorporated herein by reference in their entirety.

Background

Proper patient care requires the determination of multiple patient state metrics, which are typically measured separately using separate devices. The measured patient state metric may be as simple as pulse rate or may be more complex, such as patient temperature or blood pressure. More complex patient state metrics also include respiratory volume or blood volume. While various devices and techniques exist for measuring various patient state metrics, there is no comprehensive means of automatically monitoring these various patient metrics. Additionally, some important patient characteristics are not typically measured, but rather are qualitatively assessed by human observation. Such unmeasured patient characteristics include patient gait, lameness, body positioning, movement, fall, or ambulatory instability. Using both separate measurement devices and relying on human observation increases system complexity, reduces reliability, and increases cost.

Blood volume metrics are of particular interest due to the complexity of their measurement techniques. Conventional methods of establishing blood volume and related metrics indicative of a patient's condition have relied on highly invasive measurements of central venous pressure ("CVP" herein) or other invasive measures, such as Swan-Ganz catheterization. Such invasive measurements require special insertion of a catheter for measuring blood pressure within the central portion of the patient's circulatory system. In addition to being highly invasive, catheterization for pressure monitoring alone increases the complexity of the treatment and increases the risk of complications, such as infection. Additionally, CVP measurements can change slowly in response to certain acute conditions when the circulatory system attempts to compensate for blood volume imbalances (especially hypovolemia) by preserving blood volume levels in the central circulatory system at the expense of the periphery. For example, constriction of peripheral blood vessels can reduce the effect of fluid loss on the central system, thereby masking blood loss over a period of time in conventional CVP measurements. Such masking can lead to delayed identification and treatment of the patient's condition, leading to worse patient outcomes.

To address the problems associated with CVP measurement, the use of peripheral intravenous analysis (herein "PIVA") has been developed, as described in U.S. patent application No.14/853,504 (filed 9/14/2015 and issued as U.S. patent publication No. 2016/0073959) and PCT application No. PCT/US16/16420 (filed 3/2/2016 and issued as WO 2016/126856). Such PIVA technology uses an intravenous (herein "IV") line, such as an IV tube attached to a saline drip or IV pump, to measure peripheral venous pressure (herein "PVP"). In addition to utilizing existing IV lines, the PIVA technique includes transforming the PVP measurements into the frequency domain to identify a respiration rate frequency (F) equal to the patient's respiration rate₀) And a heart rate frequency (F) equal to the heart rate of the patient₁). While the previously disclosed PIVA techniques provide excellent indications of heart rate and blood volume status in certain situations, the previously disclosed PIVA techniques disclosed herein are further improved to address challenges associated with other situations, improve accuracy, provide early warning of potential problems, or identify additional patient conditions. Similar problems arise in other conventional methods, such as pulmonary artery or capillary pressure measurements.

Monitoring patient metrics during dialysis or other pumping presents particular challenges to both the conventional and PIVA approaches. In particular, pumping blood into the patient's circulatory system generates high levels of noise (caused by pressure variations) related to the pumping cycle. The measured signal values associated with such noise during the pumping period may be several orders of magnitude greater than the signal values associated with the non-pumping period. The prior art for monitoring patient metrics under such conditions involves shutting down the pump for long periods of time or attempting to remove the main effects of the pump from the measured pressure. Shutting down the pump for long periods of time during treatment may not be feasible in situations where continuous pumping is required, such as during surgery. Even where feasible, such methods may still result in considerable delays in determining the patient's state due to the need to interrupt pumping to obtain measurements. Similarly, prior art attempts to remove the primary effect of the pump only address the primary artifact introduced by the pump and are sensitive to errors in the estimation of the primary effect of the pump. Such techniques also typically require a priori information about the operation of the pump (e.g., the amplitude and frequency of the pressure wave generated by the pump), and some such techniques also require additional information about the precise timing of the various phases of the pump cycle. Such techniques produce only rough estimates of pressure that are not appropriate for PIVA or other advanced measures of patient condition. In particular, such techniques remove at best only an approximation of the primary artifact of the pump operation, while leaving many secondary artifacts in the measured pressure signal. Furthermore, such techniques depend on accurate estimates of the primary pumping artifacts and are sensitive to any errors in the estimates (such as errors caused by changes in pump operation over time). The technology described herein represents a means to avoid the respective problems of both types of prior art.

Accordingly, there is a need for systems and methods for filtering medical device noise artifacts from venous waveform signals.

Disclosure of Invention

In view of the disclosure herein, and without limiting the scope of the invention in any way, in a first aspect of the disclosure that may be combined with any other aspect enumerated herein unless otherwise specified, a system for monitoring a patient while the patient's circulatory system is connected to a pump using measurements associated with Peripheral Venous Pressure (PVP) within peripheral veins of the patient's circulatory system includes a PVP sensor and an evaluation unit. The PVP sensor includes a transducer disposed adjacent to or connected to an Intravenous (IV) tube in fluid connection with the peripheral vein. The PVP sensor is configured to generate an electronic signal related to the PVP while the circulatory system of the patient is connected to the pump. The evaluation unit includes a computer processor communicatively connected to the PVP sensor to receive the electronic signal and a memory storing non-transitory computer-readable instructions that, when executed by the computer processor, cause the evaluation unit to: obtaining a time-domain PVP signal comprising values of an electronic signal associated with the PVP from the transducer based on a physical phenomenon associated with the PVP of the patient over a sampling period. The sampling period includes a plurality of time segments including (i) one or more active time segments in which the pump is operating and (ii) one or more idle time segments in which the pump is not operating. The evaluation unit identifies a first plurality of values of the time domain PVP signal associated with the one or more idle time segments and a second plurality of values of the time domain PVP signal associated with the one or more active time segments based on the evaluation of the values of the time domain PVP signal. The evaluation unit generates a filtered time-domain PVP signal based on the first plurality of values and excluding the second plurality of values. The evaluation unit applies a transform to the filtered time-domain PVP signal to generate a frequency-domain PVP signal. The evaluation unit determines a patient state metric for the patient based on the frequency domain PVP signal.

In a second aspect of the disclosure that may be combined with any other aspect enumerated herein, unless otherwise specified, the pump is a peristaltic IV pump.

In a third aspect of the disclosure that may be combined with any other aspect listed herein unless otherwise specified, the pump is configured to periodically operate such that the one or more active time segments and the one or more idle time segments periodically alternate.

In a fourth aspect of the present disclosure that may be combined with any other aspect enumerated herein, unless otherwise specified, the IV tube is disposed between the patient and the pump such that a portion of the pump is in fluid connection with the peripheral vein of the circulatory system of the patient via the IV tube.

In a fifth aspect of the present disclosure that may be combined with any other aspect enumerated herein, unless otherwise specified, the transducer comprises a pressure sensor arranged in fluid connection with the interior of the IV tube, and the physical phenomenon associated with the PVP is a pressure within the interior of the IV tube.

In a sixth aspect of the disclosure that may be combined with any other aspect listed herein unless otherwise specified, the executable instructions further cause the evaluation unit to further determine whether the patient state metric indicates that the condition of the patient is abnormal, and adjust operation of the pump by changing a rate of flow of fluid from the pump into the circulatory system of the patient when the patient state metric indicates that the condition of the patient is abnormal.

In a seventh aspect of the disclosure that may be combined with any other aspect listed herein unless otherwise specified, the executable instructions that cause the evaluation unit to generate the filtered time-domain PVP signal include instructions that cause the evaluation unit to remove the one or more activity time segments from the time-domain PVP signal.

In an eighth aspect of the disclosure that may be combined with any other aspect listed herein unless otherwise specified, the executable instructions further cause the evaluation unit to generate the filtered time-domain PVP signal by, for each of the one or more pairs of the active time segment: identifying one or more corresponding values within two of the pair of active time segments; and combining the active time segments of the pair by aligning the one or more corresponding values within both of the active time segments of the pair.

In a ninth aspect of the present disclosure that may be combined with any other aspect listed herein unless otherwise specified, the executable instructions that cause the evaluation unit to generate the filtered time-domain PVP signal comprise instructions that cause the evaluation unit to: estimating a third plurality of values as replacement values for the one or more active time segments, wherein the third plurality of values are estimated based on the first plurality of values without reference to the second plurality of values. The executable instructions further cause the evaluation unit to generate the filtered time-domain PVP signal by combining the first plurality of values for the idle time segment and the third plurality of values for the active time segment.

In a tenth aspect of the disclosure that may be combined with any other aspect enumerated herein, unless otherwise specified, the third plurality of values is estimated by performing at least one of a regression analysis, a forward-backward slope calculation, a bilateral slope detection, and a mirror-matched filtering on at least the first plurality of values.

In an eleventh aspect of the disclosure that may be combined with any other aspect listed herein unless otherwise specified, the executable instructions that cause the evaluation unit to determine the patient state metric comprise instructions that cause the evaluation unit to: identifying a plurality of frequencies associated with local maxima of the frequency-domain PVP signal; and determine the patient state metric based at least in part on at least one of the plurality of frequencies associated with the local maxima.

In a twelfth aspect of the disclosure that may be combined with any other aspect recited herein unless otherwise specified, the patient state metric is a blood volume metric indicative of one or more of: hypovolemia, hypervolemia, or normovolemia.

In a thirteenth aspect of the disclosure that may be combined with any other aspect enumerated herein, unless otherwise specified, an apparatus for monitoring a patient includes a Peripheral Venous Pressure (PVP) sensor and an evaluation unit. The PVP sensor comprises a transducer configured to monitor a physical phenomenon associated with PVP within a peripheral vein of a circulatory system of a patient while the circulatory system of the patient is connected to a pump. The evaluation unit includes a computer processor communicatively connected to the PVP sensor and a memory storing non-transitory executable instructions that, when executed by the computer processor, cause the evaluation unit to: obtaining a time-domain PVP signal comprising values of an electronic signal associated with the PVP received from a transducer of the PVP sensor over a sampling period. The sampling period includes a plurality of time segments including (i) one or more active time segments in which the pump is operating and (ii) one or more idle time segments in which the pump is not operating. The evaluation unit identifies a first plurality of values of the time domain PVP signal associated with the one or more idle time segments and a second plurality of values of the time domain PVP signal associated with the one or more active time segments based on the evaluation of the values of the time domain PVP signal. The evaluation unit generates a filtered time-domain PVP signal based on the first plurality of values and excluding the second plurality of values. The evaluation unit applies a transform to the filtered time-domain PVP signal to generate a frequency-domain PVP signal. The evaluation unit determines a patient state metric for the patient based on the frequency domain PVP signal.

In a fourteenth aspect of the present disclosure that may be combined with any of the other aspects enumerated herein, unless otherwise specified, the time-domain PVP signal comprises a first time series of discrete values, the filtered time-domain PVP signal comprises a second time series of discrete values, and the second time series contains a sequential plurality of values of an order of at least one segment within the second time series, the sequential plurality of values of the order of at least one segment within the second time series corresponding to a corresponding sequential plurality of values of a corresponding segment within the first time series.

In a fifteenth aspect of the present disclosure that may be combined with any other aspect listed herein unless otherwise specified, the executable instructions that cause the evaluation unit to generate the filtered time-domain PVP signal comprise instructions that cause the evaluation unit to remove the one or more active time segments from the time-domain PVP signal.

In a sixteenth aspect of the disclosure that may be combined with any other aspect listed herein unless otherwise specified, the executable instructions that cause the evaluation unit to generate the filtered time-domain PVP signal comprise instructions that cause the evaluation unit to: estimating a third plurality of values as superseding values for the one or more active time segments, wherein the third plurality of values is estimated based on the first plurality of values without reference to the second plurality of values, and the filtered time-domain PVP signal is generated by combining the first plurality of values for the idle time segment and the third plurality of values for the active time segment.

In a seventeenth aspect of the present disclosure that may be combined with any other aspect enumerated herein, unless otherwise specified, a method of monitoring a patient while the patient's circulatory system is connected to a pump using measurements associated with Peripheral Venous Pressure (PVP) within peripheral veins of the patient's circulatory system comprises: monitoring, by a transducer, a physical phenomenon associated with the patient's PVP over a sampling period, wherein the sampling period comprises a plurality of time segments including (i) one or more active time segments during which the pump is operating and ii) one or more idle time segments during which the pump is not operating. The method comprises the following steps: obtaining, by a processor of an evaluation unit, a time-domain PVP signal comprising values of an electronic signal associated with the PVP from the transducer based on the monitored physical phenomenon over the sampling period. The method comprises the following steps: identifying, by the processor of the evaluation unit, a first plurality of values of the time-domain PVP signal associated with the one or more idle time segments and a second plurality of values of the time-domain PVP signal associated with the one or more active time segments based on the evaluation of the values of the time-domain PVP signal. The method comprises the following steps: generating, by the processor of the evaluation unit, a filtered time-domain PVP signal based on the first plurality of values and excluding the second plurality of values. The method comprises the following steps: applying, by the processor of the evaluation unit, a transform to the filtered time-domain PVP signal to generate a frequency-domain PVP signal. The method comprises the following steps: determining, by the processor of the evaluation unit, a patient state metric for the patient based on the frequency-domain PVP signal.

In an eighteenth aspect of the present disclosure that may be combined with any other aspect listed herein, unless otherwise specified, generating the filtered time-domain PVP signal comprises removing the one or more active time segments from the time-domain PVP signal.

In a nineteenth aspect of the present disclosure that may be combined with any other aspect recited herein, unless otherwise specified, generating the filtered time-domain PVP signal comprises: estimating a third plurality of values as replacement values for the one or more active time segments, wherein the third plurality of values are estimated based on the first plurality of values without reference to the second plurality of values; and generating the filtered time-domain PVP signal by combining the first plurality of values for the idle time segment and the third plurality of values for the active time segment.

In a twentieth aspect of the disclosure that may be combined with any other aspect recited herein unless otherwise specified, the third plurality of values is estimated by performing at least one of regression analysis, forward-backward slope calculation, double-sided slope detection, and image matched filtering on at least the first plurality of values.

Additional features and advantages of the disclosed apparatus, systems, and methods are described in, and will be apparent from, the following detailed description and figures. The features and advantages described herein are not all-inclusive and, in particular, many additional features and advantages will be apparent to one of ordinary skill in the art in view of the drawings and description. In addition, any particular embodiment need not necessarily have all of the advantages listed herein. Moreover, it should be noted that the language used in the specification has been principally selected for readability and instructional purposes, and may not have been selected to delineate the scope of the inventive subject matter.

Drawings

Understanding that the drawings depict only typical embodiments of the invention and are not therefore to be considered to be limiting of its scope, the disclosure will be described and explained with additional specificity and detail through the use of the accompanying drawings. The figures are listed below.

Fig. 1A illustrates a block diagram of an exemplary PIVA system for use in measuring, analyzing, and responding to a patient's peripheral venous blood pressure, the system having a fluid source.

Fig. 1B illustrates a block diagram of an exemplary PIVA system for use in measuring, analyzing, and responding to a patient's peripheral venous blood pressure, the system having no fluid source.

Fig. 1C illustrates a block diagram of an exemplary PIVA system for use in measuring, analyzing, and responding to a patient's peripheral venous blood pressure, the system including a sensor disposed within a peripheral vein.

Fig. 1D illustrates a block diagram of an exemplary PIVA system for use in measuring, analyzing, and responding to a patient's peripheral venous blood pressure, the system including a pump.

Fig. 1E illustrates a block diagram of an exemplary PIVA system for use in measuring, analyzing, and responding to a patient's peripheral venous blood pressure, the system including additional sensors for monitoring patient positioning or movement.

Fig. 2A illustrates a block diagram of an exemplary PIVA apparatus for implementing some of the functions of an exemplary PIVA system, showing fluid connections via a branch line of an IV line.

Fig. 2B illustrates a block diagram of an exemplary PIVA apparatus for performing some functions of an exemplary PIVA system, showing fluid connections via capped IV tubing.

Fig. 2C illustrates a block diagram of an exemplary PIVA apparatus for implementing some of the functions of an exemplary PIVA system, showing sensors disposed adjacent to an outer wall of an IV tube.

Fig. 3 illustrates a flow chart of an exemplary PIVA measurement and analysis method for measuring and analyzing peripheral venous blood pressure of a patient.

Fig. 4A illustrates an exemplary graph of a time domain representation of a PVP signal.

Fig. 4B illustrates an exemplary graph of a frequency domain representation of a PVP signal.

Fig. 5A illustrates an exemplary graph of a time domain representation of a PVP signal during operation of a noise-producing medical device.

Fig. 5B illustrates an exemplary graph of a time domain representation of a PVP signal after removal of an active time segment in which the medical device is operating.

Fig. 5C illustrates an exemplary graph of a time-domain representation of a filtered PVP signal including an estimate of the value for the removed active time segment.

Fig. 6 illustrates a flow chart of an exemplary pressure signal filtering method for removing noise artifacts related to the operation of a medical device from a signal corresponding to a peripheral venous blood pressure of a patient.

Fig. 7 illustrates an exemplary PIVA comparison method for identifying patient state changes based on comparison of PVPs over time.

Fig. 8 illustrates a block diagram of exemplary processing performed by an exemplary PIVA module.

Fig. 9 illustrates a block diagram of an exemplary PIVA system including a PIVA module.

Fig. 10 illustrates a block diagram of exemplary processing performed by an exemplary PIVA module.

Fig. 11 illustrates a flow chart of an exemplary patient monitoring method using a patient PVP.

Detailed Description

Although the following text sets forth a detailed description of numerous different embodiments, it should be understood that the legal scope of the invention is defined by the words of the claims set forth at the end of this patent. The detailed description is to be construed as exemplary only and does not describe every possible embodiment since describing every possible embodiment would be impractical, if not impossible. Those of ordinary skill in the art will be able to affect many alternative embodiments that will still fall within the scope of the claims. Unless a term is expressly defined herein using the sentence "as used herein, the term '_' is defined herein to mean …" or a similar sentence, there is no intent to limit the meaning of that term beyond its plain or ordinary meaning. To the extent that any term is referred to in this patent in a manner consistent with a single meaning, that is done for sake of clarity only, and it is not intended that such claim term be limited to that single meaning. Finally, unless a claim element is defined by reciting the word "means" and reciting a function without reciting any structure, it is not intended that the scope of any claim element be construed based on the application of 35 clause 112(f) of the U.S. code.

In many situations, it is important to monitor various information associated with a patient state or condition. The systems and methods disclosed herein improve on the prior art by using a metric or representation of PVP measurements to generate a patient state metric. Such a metric or representation may be generated using frequency domain PVP data derived from a time domain PVP signal corresponding to the PVP measurements. The patient state metrics may be generated using a PIVA or other similar system to monitor and respond to changes in patient condition, as further described herein. The systems, devices and methods disclosed below enable more efficient and effective monitoring by using PVP measurements to determine patient state metrics. This facilitates metric-based monitoring for a broader range of patient conditions previously susceptible to automated monitoring. This also facilitates monitoring different types of patient conditions based on measurements indicative of pressure in the peripheral vein without requiring specialized sensors to monitor each type of patient condition. Exemplary embodiments are described below.

PIVA system and signal noise

Fig. 1A-1E illustrate block diagrams of embodiments of an exemplary PIVA system 100 for use in measuring, analyzing, and responding to peripheral venous blood pressure of a patient 102. The exemplary PIVA system 100 or similar system may be used to implement various techniques for monitoring a patient's condition based on measurements associated with the PVP for the patient 102. As described below, the PIVA system 100 can measure a pressure signal associated with a peripheral vein of a patient, analyze the pressure using a PIVA technique to identify key frequency components of the pressure signal, and analyze the key frequency components of the pressure signal to determine a patient state based on one or more metrics.

The exemplary PIVA system 100 illustrated in fig. 1A includes an IV line 104 fluidly connected to the circulatory system of a patient 102. In particular, the venous access device 106 may be inserted into a peripheral vein 108 of the patient 102 at an access point. The venous access device 106 may include a needle, catheter, cannula, or other means of establishing a fluid connection between the IV tubing 104 and the peripheral vein 108. The venous access device 106 may be a separate component connected to the IV tubing 104 or may be formed as an integral part of the IV tubing 104. In either case, the venous access device 106 may include a terminating end (terminal end) that is inserted into the peripheral vein 108 at the access point and a connecting end that connects to the main portion of the IV tubing 104. A substantial portion of the IV tubing 104 may be used as a conduit between the venous access device 106 and the fluid source 110.

At some point along the main portion of the IV tubing 104, a pressure sensor 112 may be disposed to monitor physical phenomena associated with the PVP of the patient 102. In some embodiments, the pressure sensor 112 may directly measure the pressure corresponding to the PVP, such as the pressure in the interior of the IV tube 104. In such embodiments, the measurement portion of the pressure transducer (e.g., a piezoelectric pressure transducer) may be disposed in fluid connection with the interior of the IV tubing 104. The pressure sensor 112 may thus also be fluidly connected to the patient's peripheral vein 108 via the IV line 104 and the venous access device 106. Thereby enabling the pressure sensor 112 to measure pressure changes in the peripheral venous system of the patient 102 based on fluid pressure changes within the IV line 104. In other embodiments, the pressure sensor 112 may indirectly measure the pressure corresponding to the PVP of the patient 102 by measuring other phenomena without being disposed in fluid connection with the interior of the IV tubing 104. For example, the pressure sensor 112 may alternatively be attached to the outside of the IV tubing 104, thereby being fluidly disconnected from the inside of the IV tubing 104 or the fluid source 110 (as illustrated in fig. 2C). In some such embodiments, the pressure sensor 112 may measure pressure based on acoustic or optical phenomena at the sensor location. In some embodiments, the pressure sensor 112 may be disposed at a terminating end (i.e., a decapped end) of the IV tubing 104 that is specifically inserted for measuring pressure within the peripheral vein 108, in a manner similar to that illustrated in fig. 1B. In further embodiments, other sensors may be used in place of pressure sensor 112, such as acoustic, electrical, temperature, or similar sensors for measuring one or more of the following physical phenomena: pressure, sound, resistivity or conductivity, voltage or current, light level or property (e.g., spectral or frequency shift), or other similar phenomena. Regardless of which type of sensor is used, the sensor may (but need not) be in fluid contact with the patient's peripheral vein 108 through the IV tubing 104 and the venous access device 106 (or directly through the venous access device 106) to measure phenomena associated with the PVP of the patient 102. In still further embodiments, the sensor 112 may be disposed within a portion of a needle, catheter, or other venous access device 106 inserted within the peripheral vein 108 of the patient 106, as illustrated in fig. 1C. Thus, PVP can be measured in situ within peripheral vein 108. Such an in situ measurement is advantageous because it eliminates the effects of temperature, viscosity, and other factors on the transfer of pressure within the IV line 104.

In various embodiments, the pressure sensor 112 may be positioned at various distances from the access point of the peripheral vein 108, from a location within the peripheral vein 108 or proximate to the connection end of the venous access device 106, to a location proximate to the fluid source 110 or at the terminating end of the IV tubing 104. The pressure sensor 112 is illustrated in fig. 1A as being located at an intermediate location along the length of the IV tubing 104 in order to better illustrate various components of the PIVA system 100. In some embodiments, the pressure sensor 112 may directly measure the fluid pressure within the IV line 104. In particular, pressure sensor 112 may include a transducer that provides an electronic pressure signal indicative of the pressure detected by the transducer to analysis component 114 via connection 122. The electronic pressure signal may be an analog electrical signal provided directly by the transducer or may be a pre-processed digital signal indicative of a pressure value based on the transducer's interface with the main portion of the IV tubing 104. In embodiments where the pressure sensor 112 is not fluidly connected to the IV line 104 or the peripheral vein 108, the pressure sensor 112 may nevertheless include one or more transducers to generate electronic signals associated with the PVP. For example, the pressure sensor 112 may generate an electronic pressure signal indicative of the pressure within the IV tubing 104 as a proxy for the PVP within the peripheral vein 108 using one or more microphones arranged to detect sound at the outer surface of the IV tubing 104.

The analysis component 114 is communicatively connected to the pressure sensor 112 to receive the electronic pressure signal via connection 122. The analysis component 114 may include general or special purpose processing hardware, such as a microprocessor or dedicated analysis circuitry. As shown, the analysis component 114 can include one or more units for performing a PIVA analysis. Response unit 116 may identify and control a response based on pressure data from pressure sensor 112. The response unit 116 may control the presentation of an alarm or may control the operation of the fluid source 110, such as by controlling the rate of fluid flow. To determine an appropriate response, response unit 116 may receive evaluation data from evaluation unit 118, which may include metrics determined from the electronic pressure signal. The evaluation unit 118 may obtain pressure values (or signal values directly or indirectly associated with the PVP) from the electronic pressure signal and evaluate the pressure values to determine information about the patient 102, such as a blood volume metric, a positioning metric, a movement metric, or other metrics as described in more detail below. The information generated by the evaluation unit 118 may also be stored or presented for patient monitoring. In alternative embodiments, additional, fewer, or alternative elements may be included. For example, the evaluation unit 118 may perform the functions attributed herein to the response unit 116.

In some embodiments, the analysis component 114 may be communicatively connected to the monitor 120 via connection 126. The monitor 120 may be a separate monitor for displaying information about the patient or may be incorporated into another device, such as a pump or other fluid source device. The monitor 120 may also be communicatively connected to the fluid source 110 via a connection 128 to receive and display information associated with the fluid source 110. In some embodiments, the monitor 120 may be used to control the operation of the fluid source 110, such as by adjusting the fluid flow rate, the duration of operation, the mode of operation, or other similar controls. In some embodiments, the analysis component 114 may similarly be communicatively connected to the fluid source 110 via a connection 124. The analysis component 114 can receive information regarding the operation of the fluid source 110 for use in evaluating a patient by the evaluation unit 118. The response unit 116 may also be in communication with the fluid source 110 to control operation of the fluid source 110 in response to information about the patient determined based on the electronic pressure signal from the pressure sensor 112.

In some embodiments, the fluid source 110 may include a pump 111, as illustrated in fig. 1D. Such a pump may be disposed within the exemplary PIVA system 100 to pump blood or other fluids into the peripheral vein 108 of the patient 102. For example, the pump 111 may comprise an IV infusion pump or a dialysis pump, such as a peristaltic pump. The pump 111 may be configured to operate cyclically in a periodic or aperiodic manner, having alternating operating intervals (i.e., active time segments) and rest times (i.e., idle time segments). By alternating the pump 111 between the on interval and the off interval, periods of time when the pump 111 is not on can be used for PIVA analysis, as described further below. In some embodiments, such as where the pump 111 is a hemodialysis pump, the pump 111 may also be connected to the circulatory system of the patient 102 through an additional IV tube 105 (which may include or be further attached to an additional venous access device 107), creating an extracorporeal blood circuit through the pump 104 via

tubes

104 and 105. In such embodiments, the pump 111 may draw blood from the patient 102 through either of the

tubes

104 or 105. The extracorporeal blood may then be treated according to a treatment protocol (or may be replaced with another fluid that may be infused into the patient's circulatory system) before being sent back to the patient's circulatory system through the other of the

IV tubes

105 or 104. Although described herein as one assembly, it should be understood that in some embodiments, pump 111 may include multiple pumping assemblies (e.g., a pair of pumps for extracting and returning blood or other fluids or multiple pumps in a common fluid system).

In some embodiments, the example PIVA system 100 may include one or more additional sensors 150, as illustrated in fig. 1E. The additional sensors 150 may include pressure sensors, infrared sensors, optical sensors, magnetic sensors, and the like. In various embodiments, each additional sensor 150 may be connected to the analysis component 114 via a connection 152 or to the monitor 120 via a connection 154, which may be a wired or wireless connection. Such additional sensors 150 may be arranged to monitor the presence, absence, location or position of the patient 102. For example, a pressure sensor may be disposed within a hospital bed to determine whether the patient 102 is within the bed based on weight measurements. Similarly, one or more sensors may be arranged to determine whether such a bed is flat or partly elevated for sitting. Other additional sensors 150 may be disposed on the patient 102 to monitor movement. For example, the patient 102 may wear a wrist band sensor containing an accelerometer array, which may measure data about at least some patient movement. The additional sensor 150 may thus be disposed with the pressure sensor 112 within the PIVA device 130 or may be separate from the pressure sensor 112. In further embodiments, the additional sensor 150 may further include any of the following to measure orientation or motion of the patient: a real-time three-dimensional gyroscope, one or more cameras monitoring a local physical environment surrounding the patient, or a microphone configured to monitor sound in the local physical environment. The sensor data from the additional sensors 150 may be correlated with IV pressure measurements or other pressure-related measurements associated with the patient's PVP.

In various embodiments, the

various connections

122, 124, 126, and 128 may each be wired or wireless connections. Further, some or all of the

connections

122, 124, 126, and 128 may be internal to a device such as the PIVA device 130 or the integrated PIVA fluid source 140.

The PIVA device 130 can incorporate the pressure sensor 112 and the analysis component 114 (and associated connections) into a device that can be attached to the IV line 104 or within the IV line to perform PIVA monitoring of the patient 102. In some embodiments, the PIVA device 130 can further include one or more additional sensors 150 or other components described herein. The PIVA-integrated fluid source 140 can include a computer-controlled fluid reservoir or pump configured to utilize PIVA monitoring of the patient 102 in controlling fluid flow. Like the PIVA device 130, the PIVA-integrated fluid source 140 may include the pressure sensor 112 and the analytical component 114 as well as the fluid source 110 and the monitor 120 (and associated connections). Alternative embodiments may include additional, fewer, or alternative components in alternative configurations.

Fig. 2A-2C illustrate block diagrams of exemplary embodiments of a PIVA device 130 for implementing some of the functionality of the exemplary PIVA system 100. As illustrated in fig. 2A, the example PIVA device 130 may be configured to attach to a branch line 104A of the IV tube 104, such as at one limb of a Y-connector or T-connector. Alternatively, the example PIVA device 130 may be configured to attach to the terminating end of the IV tubing 104, as illustrated in fig. 2B. In such an embodiment, the PIVA device 130 may be capped on the terminating portion of the IV tubing 104 such that no fluid source 110 is connected to the peripheral vein 108 through the same IV tubing 104. Of course, the fluid source can be connected in other ways to provide fluid to the patient 102 via another IV tube and another venous access device. In further embodiments, the PIVA device 130 can be configured to attach to the outside of the IV tubing 104, as illustrated in fig. 2C. In such embodiments, one or more sensors of the PIVA device 130 may monitor the PVP without an internal fluid connection to the peripheral vein 106 or the IV line 104.

As discussed above, the PIVA device 130 can include a pressure sensor 112, the pressure sensor 112 being disposed such that the sensing portion is in contact with the fluid in the IV tubing 104, as illustrated in fig. 2A and 2B. In some embodiments, the pressure sensor 112 (or an alternative sensor) may alternatively be external to the IV tubing 104, as illustrated in fig. 2C. Wherever located, pressure sensor 112 is arranged to monitor physical phenomena associated with pressure in peripheral vein 108. Such physical phenomena may include pressure in the IV tubing 104, expansion or contraction of the IV tubing 104, sound in the IV tubing 104, vibration of the IV tubing 104, or other similar phenomena. Pressure sensor 112 may be electrically communicatively connected to microprocessor 132 via system bus 138. The microprocessor 132(MP) may be further communicatively connected to a program memory 134 and a communication unit 136 (communication unit) via a system bus 138. Program memory 134 may be a non-transitory non-volatile memory (e.g., flash memory) that stores executable instructions that may be executed by microprocessor 132 to evaluate the electronic pressure signal from pressure sensor 112, determine patient information (e.g., blood volume metrics), determine an appropriate response to the determined patient information, and control communication unit 136 to electronically communicate with fluid source 110 or monitor 120 via

connections

124 or 126. The program memory 134 may store a plurality of routines, scripts or modules corresponding to the units or sub-units of the analysis component 114, such as software modules corresponding to the response unit 116 or the evaluation unit 118.

The communication unit 136 may be a hardware component configured to send and receive electronic data between the PIVA device 130 and the fluid source 110 or monitor 120 via the

connections

124 or 126. The

connections

124 and 126 are illustrated as wired connections in the exemplary PIVA device 130, which may also be used to obtain power for the PIVA device 130. Alternatively, another power connection or battery (not shown) may provide power to the PIVA device 130. Although shown as separate wired connections, the

connections

124 and 126 may be separate or combined wired or wireless connections. The

connections

124 and 126 may be in communication with a communication component of the fluid source 110 or the monitor 120, which fluid source 110 or monitor 120 may comprise a portion of the pump 111 or be a portion of the pump 111. Such communications may include raw data generated by the pressure sensor 112, processed data related to measurements made by the pressure sensor 112, data analyzed according to methods described below, or alarm signals or control commands determined based on the analyzed data. The fluid source 110 or monitor 120 can then take appropriate action or present appropriate information based on the communication from the example PIVA device 130.

Fig. 3 illustrates a flow chart of an exemplary PIVA measurement and analysis method 300 for measuring and analyzing the condition of the patient 102 based on PVP using the PIVA system 100. The method 300 may be used to determine various patient state metrics, such as metrics related to the patient's blood pressure, blood volume, respiration, positioning or movement, or systemic vascular resistance. The method 300 may be performed by the evaluation unit 118 using an electronic pressure signal from the pressure sensor 112, which in some embodiments may be included in the method 300 as being generated by the pressure sensor 112.

The method 300 begins by measuring a PVP data signal for the patient 102 (block 302). The PVP data signal may be measured by using the transducer of the pressure sensor 112 to generate an electronic pressure signal indicative of the PVP based on a physical phenomenon associated with the PVP. This may be accomplished, for example, by measuring the pressure within the IV line 104. Because the IV line 104 is fluidly connected to the peripheral vein 108 of the patient 102 via the venous access device 106, the pressure in the IV line 104 measured by the pressure sensor 112 is associated with the patient PVP (i.e., the pressure in the peripheral vein 108). In some embodiments of the PIVA system 100, the pressure within the IV line 104 may be different from the PVP within the peripheral vein 108, but the pressure measured within the IV line 104 may still be proportional to the PVP within the peripheral vein 108. Thus, the measured PVP data signal may be adjusted as needed to compensate for differences between pressures. For example, adjustments may be made based on temperature, viscosity of the patient's blood or fluid provided by the fluid source 110, or the gauge or rigidity of the IV tubing 104. The PVP data signal measured by the pressure sensor 112, whether adjusted or unadjusted, accurately represents changes in pressure over time, including periodic pressure changes associated with respiration and circulation periods, as well as aperiodic pressure changes that may indicate changes in patient condition. Similarly, the PVP data signal generated by the pressure sensor 112 through components not in contact with the internal fluid of the IV tube 104 likewise provides a representation of the pressure within the peripheral vein 108 of the patient 102. The PVP data signal may be an electronic pressure signal generated by the pressure sensor 112 or may be a data signal derived from the electronic pressure signal. In alternative embodiments, the PVP data signal may be evaluated in real time as it is generated, or it may be stored for later analysis. Depending on the component used to measure PVP-related phenomena, the PVP data signal may be generated or stored as an analog signal (i.e., as a continuous function or curve over some time segment) or a digital signal (i.e., as a set of discrete values representing different times).

Fig. 4A illustrates an exemplary graph of a time domain representation of the PVP data signal, which may be an electronic pressure signal from the pressure sensor 112. The graph illustrates a time-domain PVP signal 402, which time-domain PVP signal 402 shows periodic increases and decreases in pressure associated with the patient's heartbeat. Additionally, the time-domain PVP signal 402 exhibits a relatively slow periodic variation as a result of the patient's breathing. The graph also illustrates a respiration curve 404, which respiration curve 404 shows the effect of inspiration and expiration on the time-domain PVP signal 402. As the lungs expand during inspiration, the pressure measured in the peripheral veins is higher during inspiration than during expiration as the volume of the lungs decreases. Other factors such as blood volume and patient movement affect PVP.

The time-domain PVP signal 402 is thus a combination of various effects, periodic (e.g., heart rate or respiration) and aperiodic (e.g., movement or blood loss). Because the resulting time-domain PVP signal 402 will include noise from various sources, it may be difficult to detect small pressure changes that may be used as an indication of the patient's state. Thus, as described below, in some embodiments, the PIVA technique utilizes frequency domain evaluation of the PVP data signal. In other embodiments, time domain or hybrid techniques may also be used to assess patient status or generate patient status metrics. It should be appreciated that while the time domain representation of the PVP data signal is graphically illustrated in fig. 4A as a graph to illustrate the salient features of the data, it is not necessary to produce a graph or other graphical representation of such a data signal. Alternatively, in some embodiments, the PVP data signal is processed by the evaluation unit 118 without generating a graphical representation of the time-domain PVP data signal, or the graphical representation may be generated separately from the evaluation for review by the user.

Returning to fig. 3, a plurality of data values may then be obtained from the measured PVP data signal (block 304). The evaluation unit 118 may sample values of the live or stored PVP data signal to obtain the plurality of data values. In some embodiments, the data values may be sampled at fixed intervals over a period of time to obtain a plurality of data values within an evaluation window, which may include storing the plurality of data values associated with the window in a temporary or permanent electronic data storage device. In further embodiments, data for multiple evaluation windows may be obtained such that each evaluation window includes multiple data values. For example, concurrent time periods may be identified as separate evaluation windows, or evaluation windows may be identified as time periods separated by intervening time periods (e.g., twenty-two evaluation windows beginning every minute, thus separated by forty-two intervening time periods). When the evaluation unit 118 samples values of a live (continuously updated) PVP data signal, in some embodiments the evaluation window may be updated on a rolling basis to obtain new data values while covering a fixed duration period of time. For example, the evaluation window may be repeatedly updated to maintain a window of fixed duration (e.g., five seconds, ten seconds, twenty seconds, or some other period of time) of the most recent PVP data from the pressure sensor 112 by adding new sample data values and removing the oldest sample data values. Where the evaluation unit 118 periodically obtains updates of new sample data values, the window may be updated each time a new data value is received (and the transformations and evaluations described below may be performed on the updated window). In an alternative embodiment, the plurality of data values may correspond to continuous values of an analog PVP data signal, which may be obtained and analyzed by analog electronics (which may be part of the evaluation unit 118).

From the plurality of data values, evaluation unit 118 generates frequency domain data corresponding to the plurality of data values (block 306). Such frequency domain data may be generated as a frequency distribution representing the PVP data signal in the frequency domain as an amplitude associated with each of a plurality of frequencies. This may include applying a data transform to the plurality of data values representing the time-domain PVP signal to produce a frequency-domain representation of the PVP signal. In a preferred embodiment, the evaluation unit 118 applies a Fast Fourier Transform (FFT) to the sampled plurality of data values to generate a frequency domain representation of the PVP signal. In different embodiments, different data transforms (e.g., laplace transform, mellin transform, hartley transform, short-time fourier transform, chirped wavelet transform, hankel transform, or any other continuous or discrete transform) may be implemented to transform the data into a frequency domain representation of the PVP signal. The FFT may be applied periodically (e.g., every ten seconds, every minute, or every two seconds, with or without overlapping evaluation windows). In some embodiments, other analysis techniques are contemplated that can identify local maxima as a function of frequency, such as wavelet transforms, autocorrelations, or other signal analysis techniques that can separate contributions to the signal spectral energy content over time-domain segments.

The frequency domain data may comprise a plurality of values representing the amplitudes of various frequency components in the measured PVP data signal based on the plurality of data values. Such values may be discrete or may be part of a curve of amplitudes corresponding to frequencies, which may be generated by interpolation or approximation between a finite number of values associated with a finite number of frequencies. Although an FFT algorithm may be used to produce a large effect, the plurality of data values may be evaluated using other time-frequency transforms or other techniques that analyze the frequency components of the signal. For example, the evaluation may include a wavelet transform or time-frequency representation of the measured PVP data signal, in addition to other fourier transforms.

Fig. 4B illustrates an exemplary graph of a frequency domain representation of the PVP data signal corresponding to the time domain PVP signal 402 represented in the time domain in fig. 4A. The graph illustrates the amplitude of each frequency component by a frequency curve 406. Typically, the horizontal axis represents frequency and the vertical axis represents amplitude. Although a chart is exemplary, certain characteristic features may be discerned in the chart. Of particular interest are the sum frequencies (F)_N) Several peaks (P) of the associated frequency curve 406_N). Between peaks, a small change in amplitude is seen, which may represent a small component of the time domain PVP signal 402 associated with noise in the system or artifacts of the circulatory system of the patient 102 (e.g., patient movement during measurement, or opening and closing of the atrioventricular and aortic valves) or in the exemplary PIVA system 100 (e.g., pump noise).

While the frequency domain representation of the PVP data signal is illustrated as a graph in fig. 4B to illustrate the salient features, it should be understood that it is not necessary to produce a graph or other graphical representation of the frequency domain data. Indeed, in some embodiments, no such graphical representation is generated. Alternatively, the frequency domain data is processed by the evaluation unit 118 as an intermediate process, the results of which are not presented directly to the user of the system or device. In some embodiments, the frequency domain data may be stored in a temporary or non-temporary memory as values within a data list, data table, or similar data structure.

In the normal case, with the lowest frequency (F)₀) Peak value (P) of₀) Corresponding to the respiratory rate of the patient 102, but with the next lowest frequency (F)₁) Peak value (P) of₁) Corresponding to the heart rate of the patient 102. In some embodiments, a heart rate frequency (F) may be identified₁) Harmonic frequency (F) of_H) Associated one or more harmonic peaks (P)_H). Such harmonic peak (P)_H) Associated with a local maximum of the frequency curve 406. The next two peaks (P) of the frequency curve 406₂) And (P)₃) Is at the second harmonic frequency (F) with the heart rate₂) And the second harmonic frequency (F)₃) Harmonic peaks (P) occurring at frequencies associated with the next first and second harmonics_H). Harmonic generation at heart rate frequency (F)₁) At a fixed multiple of. Typically, these multiples are typically integer multiples. In particular, experimental data indicate the first harmonic frequency (F)₂) About heart rate frequency (F)₁) Twice of the second harmonic frequency (F)₃) About heart rate frequency (F)₁) Three times that of the original.

Such as for corresponding frequencies (e.g., F) via the evaluation unit 118₁、F₂、F₃) Peak value of (e.g., P)₁、P₂、P₃) Provides for subsequent calculation of the patient's state (e.g., hemodynamic state). E.g. corresponding frequency (e.g. F)₁、F₂、F₃) Peak value of (e.g., P)₁、P₂、P₃) Can be used to calculate the PIVA score, as further detailed herein.

Although not shown, howeverAdditional peaks associated with third and higher harmonics of the heart rate may be identified in some embodiments. Usually at heart rate frequency (F)₁) At a corresponding sequential integer multiple of (F) occurs a further harmonic frequency (F)₄、F₅、…F_N). For example, can be passed through F₃To represent the second harmonic frequency, which can be represented by F₄To indicate third harmonic frequencies, etc. Although there is some variation in the frequency of the peaks observed to be associated with harmonic frequencies, it has been found that these are generally located above or below the heart rate frequency (F)₁) The harmonic frequency peaks occur within about ten percent (i.e., ± 10%) of the heart rate frequency values of the integer multiple. Peak value (P)_N) May vary, but is related to the heart rate frequency (F)₁) Associated peak value (P)₁) Should be greater than the harmonic frequency (F) thereof₂)、(F₃) Etc. associated peak value (P)₂)、(P₃) Etc.

Further, it should be noted that although fig. 4B illustrates the frequency curve 406 as a number of parabolic peaks (e.g., P₀、P₁、P₂、P₃) Other graphical representations of the frequency domain representation are contemplated. For example, to the extent that the system is consistent (e.g., consistent patient breathing and heart rate) and the sampling rate is sufficiently high (e.g., the sampling rate of data values measured in the time domain), the peak (e.g., P) may be adjusted to account for the peak (e.g., P) and the heart rate₀、P₁、P₂、P₃) Depicted diagrammatically as vertical lines (e.g., parabolic peaks with no appreciable width or parabolic peaks without width).

Although the present disclosure generally refers to the respiration rate as corresponding to the lowest frequency peak (P)₀) The heart rate is said to correspond to the next lowest frequency peak (P)₁) And so on, it should be appreciated that any such reference is made for ease of illustration. To this end, in some embodiments, the time-domain PVP signal may detect one or more frequencies below the respiration rate. For example, gut frequencies tend to be associated with lower than typical respiratory frequencies. In these embodiments, there is the lowest frequency (F)₀) Peak value (P) of₀) Corresponding to the intestinal frequency but with the next lowest frequency (F)₁) Peak value (P) of₁) Corresponding to the breathing frequency. Similarly, each of the heart rate frequency and the corresponding harmonic frequency will correspond to the next lowest peak (P), respectively₂) And the following peak (P)₃、P₄、...P_N). It will be appreciated that in some further embodiments, the time-domain PVP signal may detect a plurality of frequencies below the breathing frequency. Thus, the peak indices corresponding to the respiration rate, heart rate, and heart rate harmonics may increase the number of frequencies detected as being below the respiration rate. Therefore, unless specifically described otherwise, the peak value (P) is correlated with the lowest frequency₀) Corresponding respiration rate and the next lowest frequency peak (P)₁) Any reference to a corresponding heart rate frequency is not limiting, and it is also contemplated to shift the corresponding peak index by the number of frequencies below the respiration rate detected by the time-domain PVP signal.

Turning again to fig. 3, the evaluation unit 118 further identifies a peak (P) associated with the frequency domain representation of the PVP signal_N) Corresponding multiple frequencies (F)_N) (block 308), such as frequency curve 406. The evaluation unit 118 may first identify a peak (P) in the frequency domain representation indicative of the PVP signal by comparing the frequency domain PVP signal values_N) And then identifying the peak value (P) associated with the identified peak value_N) Associated corresponding frequency (F)_N). To determine the peak value (P)_N) The evaluation unit 118 may identify local maxima as peaks using any of a variety of methods, including methods based on any or all of a comparison of the relative magnitudes of the local maxima, the establishment of a fixed or dynamic frequency band around each peak, or a comparison of the full width at half maximum for the local maxima. For example, a band pass filter may be employed to separate segments of the frequency domain representation of the PVP signal to further identify local maxima. This is in identifying the harmonic peak (P)_N) And corresponding harmonic frequency (F)_H) It may be particularly useful because such harmonics occur at the heart rate frequency (F)₁) At integer multiples of.

As an example, centering at heart rate frequency (F) may be used₁) About twice the frequency and has a heart rate frequency (F)₁) A band pass filter having a bandwidth of twenty percent to define a band pass filter containing a first harmonic peak (P)₂) Is the range of the frequency domain representation of the PVP signal of. The first harmonic frequency (F) may then be identified by simply determining the frequency associated with the local maximum of the frequency domain representation of the PVP signal in such a range₂). By employing these or other known techniques, the peak (P) of the frequency domain representation of the PVP signal can be represented_N) Distinguished from other local maxima caused by noise or other minor phenomena in the circulatory system.

Once and peak (P)_N) Associated with said plurality of frequencies (F)_N) Having been identified, the evaluation unit 118 may analyze the frequency domain representation of the PVP signal at one or more frequencies (F)_N) To determine one or more aspects of the patient state (block 310). Such analysis may include determining one or more patient state metrics, such as a blood volume metric, a respiratory volume metric, a patient positioning metric, a patient movement metric, a systemic vascular resistance metric, other metrics related to systemic vascular resistance (e.g., mean arterial pressure, mean venous pressure, cardiac output), and so forth for the patient 102. For example, the patient state metric may include a blood volume metric indicative of one of the following hemodynamic states of the patient 102: hypovolemia, hypervolemia, or normovolemia. In various embodiments, the hemodynamic status of the patient 102 may be determined as a score or category of patient status. In further embodiments, time domain analysis may additionally or alternatively be performed to evaluate PVP signals, as discussed elsewhere herein.

May be directly dependent on one or more frequencies (F)_N) Or the amplitude of the frequency domain representation of the PVP signal associated therewith, to determine some patient state metric. For example, it can be based on the respiratory rate (F)₀) The associated amplitude (i.e. the respiration peak (P))₀) Amplitude of) to determine depth of breath, or may be based on a correlation with heart rate frequency (F)₁) Associated amplitude (i.e. heart rate peak (P))₁) Amplitude of) to determine a blood volume metric. As another example, it may be measured directly orA blood volume metric indicative of the patient's hemodynamic status (e.g., hypovolemia or hypervolemia) is calculated.

For example, as previously mentioned, following performing the transformation, the evaluation unit 118 may identify the corresponding frequency (e.g., F)₁、F₂、F₃) Peak value of (e.g., P)₁、P₂、P₃). Various frequencies (such as heart rate frequency F) may then be used in the equation₁First harmonic of heart rate frequency F₂And second harmonic F of heart rate frequency₃) Corresponding individual peaks (e.g., P)₁、P₂、P₃) To calculate the PIVA score. The PIVA score, which represents the fluid state of the patient, is also a corollary to the pulmonary capillary wedge pressure. Because pulmonary capillary wedge pressure is an indicator of fluid status (e.g., hypovolemia or hypervolemia), the PIVA score likewise represents the fluid status of the patient.

In one embodiment, the equation for calculating the PIVA score is represented by:

c₀、c₁、c₂、c₃、g₀、g₁、g₂、g₃、h₀、h₁、h₂、h₃、i₀、i₁、i₂and i₃Each of which is a constant. mag^f1、mag^f2And mag^f3Each representing a respective frequency (e.g., F)₁、F₂、F₃) An individual amplitude of each of the. These amplitudes are also commonly referred to herein as peaks in frequency. For example, mag^f1May also be referred to herein as correlating with heart rate frequency F₁Associated peak value P₁. Similarly, for example, mag^f2May also be referred to herein as a harmonic of the first harmonic frequency F₂Associated peak value P₂. Similarly, for example, mag^f3May also be referred to herein as harmonic second harmonic frequenciesF₃Associated peak value P₃. For example, and referring to FIG. 4B, it is referred to as mag in the PIVA equation^f1P of₁Is the heart rate frequency (F)₁) Is called mag in the PIVA equation^f2P of₂Is the first harmonic frequency (F)₂) And is referred to as mag in the PIVA equation^f3P of₃Is the first harmonic frequency (F)₃) Of the amplitude of (c).

The evaluation unit 118 calculates a PIVA score, which is unitless. In a related embodiment, the PIVA system 100 displays the PIVA score (e.g., via the monitor 120). By calculating the PIVA score, the fluid status of the patient (e.g., hypovolemia, hypervolemia, or edema) can be readily determined. Preferably, the calculated PIVA score is consistent with a pulmonary capillary wedge pressure of ± 8mmHg at the limit of consistency at the 95% confidence interval.

In one embodiment, additional peak amplitudes (e.g., corresponding to F) corresponding to various frequencies may also be used in calculating the PIVA fraction (e.g., also implementing additional constants)₄P corresponding to third harmonic frequency₄) To obtain greater accuracy in the calculations.

In one embodiment, the calculation or measurement may be directly related to the harmonic frequencies (F)_H) Associated harmonic peak (P)_H) Is determined, such as is related to a change in amplitude at a previous time when the hemodynamic state of the patient is known (e.g., a pre-operative baseline measurement). As yet another example, the heart rate frequency (F) may be based₁) Variation over time or by measurement with peak heart rate (P)₁) The heart rate variability is determined by the width of a portion of the frequency domain representation of the associated PVP signal (e.g., full width at half maximum).

In some embodiments, the frequency peak (F) may be based on a difference in frequency peaks (F) based on the same plurality of data values (i.e., for the same evaluation window)_NAnd F_M) The comparison of the associated amplitudes determines a patient state metric. For example, with heart rate frequency F₁And the first harmonic frequency F₂Ratio of associated amplitudesCan be used to determine a patient's hemodynamic metric, such as systemic vascular resistance or blood volume fraction. Such a ratio is at the harmonic frequency (F)_H) The associated amplitudes may be particularly useful in normalizing to obtain a more robust and accurate patient state metric. Similarly, with a different harmonic frequency (e.g., F)₂And F₃) The ratio between the amplitudes of the frequency domain representations of the associated PVP signals may be used to determine the hemodynamic state (e.g., blood volume) of the patient 102. In further embodiments, the peak value (P) may be determined based on a plurality of data values other than for different ones (i.e., for different evaluation windows)_N) Of the same frequency or frequencies (F)_N) The comparison of the associated amplitudes determines a patient state metric. For example, for the heart rate frequency F₁Analysis of the change in the associated absolute or relative amplitude over time may be used to determine a hemodynamic metric. Information about the patient state may be stored in memory, presented to a user via monitor 120, or used by response unit 116 to generate and implement a response (e.g., to present an alarm or to control the operation of fluid source 110), including any of the responses discussed further below.

In further embodiments, additional information about the patient may be used in determining some patient status metrics, or such additional information may be monitored for use with the patient status metrics. For example, information about patient positioning or movement (e.g., patient movement metrics) may be separately monitored to provide context for patient state metrics or to supplement patient state metrics. To this end, additional patient metrics may be monitored individually by additional sensors 150 that collect data about the positioning or movement of the patient 102, or multiple patient metrics may be determined by analysis of the PVP signal monitored via the pressure sensor 112. For example, PVP measured from pressure sensor 112 (such as heart rate frequency F) may be combined₁Or associated amplitude P₁) A sudden offset in the derived patient metrics and a spike in the measured acceleration from the additional sensor 150 to determine that the patient is likely to have fallen. As another example, it may be implementedBoth frequency domain analysis of the PVP signal and time domain analysis of the PVP signal (e.g., waveform analysis or pattern detection) are performed to generate patient metrics, which may then be combined or analyzed together to assess the patient state. Additional patient metrics may be evaluated to verify the appropriateness of the response to changes in the patient state metric. Thus, if the patient state metric indicates a likelihood of an instantaneous condition at the same time as the additional patient metric indicates patient movement, the patient state metric may be determined as a result of the patient movement, so no response may be required. Alternatively, if the additional patient metrics confirm patient status metrics indicating that the patient moved or worked for patients requiring ambulatory assistance, an alarm may be generated to alert responsible personnel that the patient may be attempting to walk without assistance. In some embodiments, the additional information may include information indicative of patient condition or restrictions, such as patient condition information entered by a doctor or nurse.

Fig. 5A-5C illustrate exemplary graphs of time domain representations of PVP signals including noise artifacts, such as from the operation of the pump 111 or other fluid source 110. These exemplary diagrams illustrate various stages or types of processing that can be performed by the analysis component 114. Fig. 5A illustrates a PVP data signal 502, which PVP data signal 502 includes both an inactive segment 502I associated with an idle time segment in which the pump 111 is not operating and an active segment 502A associated with an active time segment in which the pump 111 is operating. To illustrate the effect of pump activity on the PVP data signal 502, fig. 5A further illustrates the operation of the pump 111 by plotting the pump control signal 504 on the same time scale. For simplicity, the pump control signal 504 is illustrated as a binary signal, where a value of "1" indicates active pumping and a signal of "0" indicates inactive. However, in alternative embodiments, alternative types of pump control signals may be used to control the power or operating mode of the pump 111.

As illustrated in FIG. 5A, pump 111 is at time t₀And t1, so the value of the PVP signal 502 during this time segment forms the inactive pump PVP signal 502I. The inactive pump PVP signal 502I indicates no interference from the pump 111PVP measurements corresponding to pressure in the circulatory system of the patient 102. Thus, the inactive pump PVP signal 502I is similar to the time-domain PVP signal 402 described above. Thus, as discussed herein, the value of the inactive pump PVP signal 502I may be used to perform further analysis in accordance with the PIVA or other frequency domain methods. As further indicated, pump 111 is at time t immediately after the first idle time segment₁And t₂During the first active time segment in between. The values of the PVP signal 502 during the first active period form an active pump PVP signal 502A, the values of which include noise artifacts from the operation of the pump 111. Such noise artifacts of the active time segments suppress the PIVA and other related analyses, so it is useful to remove, replace, or otherwise adjust the active pump PVP signal 502A prior to further analysis. The additional second and third idle time segments associated with the inactive pump PVP signal 502I during which the pump 111 is not operating are further illustrated at time t₂And t₃And at time t₄And t₅In the meantime. The additional active time segment associated with the active pump PVP signal 502A during which the pump 111 is operating is illustrated at time t₃And t₄In the meantime. Although active and idle time segments are illustrated in the exemplary diagram as being adjacent in time, some embodiments may include transition periods that are neither part of any idle time segment period nor part of any active time period.

Fig. 5B illustrates an exemplary cleared PVP signal 508 including only the inactive pump PVP signal 502I. The example cleaned PVP signal 508 may be generated by simply removing the data values associated with the active time segments, leaving a gap 506 in the cleaned PVP signal 508. To remove active time segments, the analysis component 114 can first identify one or more of either or both active time segments or idle time segments. In some embodiments, information from the pump 111 (such as the pump control signal 504) may be used to identify active time segments or idle time segments. However, in a preferred embodiment, the analysis component 114 may identify an active time segment or an idle time segment based on the value of the PVP signal 502. The analysis component 114 can identify active time segments or idle time segments based on the magnitude or value of the PVP signal 502 changes, as discussed further below.

Once generated, the cleaned PVP signal 508 may be analyzed directly according to the methods described herein, or the cleaned PVP signal 508 may be further adjusted prior to transformation into the frequency domain. For example, the cleared PVP signal 508 may be adjusted to remove the gap 506 by aligning the inactive pump PVP signal 502I to partially overlap based on the periodicity of the inactive pump PVP signal 502I. As another example, the cleared PVP signal 508 may be adjusted based on the inactive pump PVP signal 502I to fill the gap 506 with the estimated value, as illustrated in fig. 5C. Alternatively, rather than estimating the gap 506, the inactive pump PVP signal 502I may be connected via other means, such as via a straight line connecting the end point of one inactive pump PVP signal 502I to the start point of a second inactive pump PVP signal (e.g., a straight line spanning the gap 506). While the inactive pump PVP signal 502I associated with a single idle time segment may be sufficient for frequency domain analysis of the patient state metric if the duration of the idle time segment is long enough, the duration of the idle time segment may be too short to allow accurate analysis. In such a case, combining the plurality of inactive pump PVP signals 502I over a corresponding plurality of idle time segments facilitates further analysis by providing more data for evaluation. Even when the individual idle time segments are long enough to allow frequency analysis, accuracy may be improved by adding additional data values associated with additional idle time segments.

Fig. 5C illustrates an exemplary adjusted PVP signal 510 including the inactive pump PVP signal 502I and the estimated PVP signal 502E to fill the gap 506. The value of the estimated PVP signal 502E may be estimated based on the value of the inactive pump PVP signal 502I of the cleared PVP signal 508, as discussed further below. By filling the gap 506 with the estimated PVP signal 502E, the resulting adjusted PVP signal 510 may be better suited for some types of further analysis. In particular, the adjusted PVP signal 510 represents a comprehensive time series of data without noise artifacts from the operation of the pump 111 that can be analyzed without further adjustment for the effects of pumping. It should be noted that the adjusted PVP signal 510 may be obtained solely from the measured PVP signal 502 without reference to extrinsic data about the pump 111. Accordingly, extrinsic data regarding the time of pump operation (e.g., the period of pump operation) or characteristics of pump operation (e.g., pump speed, pump capacity, or a model of noise artifacts generated by the pump) is not required to generate the adjusted PVP signal 510.

Although fig. 5C illustrates the estimated PVP signal 502E as filling only the gap 506 created by removing the active pump PVP signal 502A, some embodiments may include estimating the entire adjusted PVP signal 510. In such embodiments, the active pump PVP signal 502A and the inactive pump PVP signal 502I may be replaced with the estimated PVP signal 502E to generate the adjusted PVP signal 510. While such a method may in some aspects reduce the accuracy of the analysis by replacing the measurement of the inactive pump PVP signal 502I with the estimated value of the estimated PVP signal 502E, the method may reduce the accuracy of the analysis by eliminating the boundary between the active time segment and the idle time segment (i.e., at time t, t)₁、t₂、t₃And t₄At) to better facilitate further analysis. In still further embodiments, the discontinuity may be addressed by adjusting the value of one or more of the inactive pump PVP signal 502I or the estimated PVP signal 502E occurring near the boundary between the active time segment and the idle time segment to smooth the transition. In any case, the active pump PVP signal 502A is excluded from the adjusted PVP signal 510 and replaced with the estimated PVP signal 502E.

Fig. 6 illustrates a flow chart of an exemplary pressure signal filtering method 600 for removing noise artifacts related to the operation of a medical device from a signal corresponding to the PVP of the patient 102. The filtering method 600 may be implemented by the evaluation unit 118 to obtain the PVP signal, filter the PVP signal and analyze the PVP signal to determine the patient state metric. Noise artifacts from the operation of the pump 111, other fluid source 110, or similar medical device may obscure normal PVP measurements during operation. For analytical methods such as PIVA, these noise artifacts must be removed or otherwise addressed before further processing in order to obtain an accurate metric. In contrast to other methods of addressing device noise artifacts, the filtering method 600 identifies and removes signal values associated with active time segments from the PVP signal containing active time segments of device operation and idle time segments when the device is inactive. To do this, a time-domain PVP signal (such as PVP signal 502) is obtained and processed to remove signal values associated with the active time segment (such as the active pump PVP signal 502A) to generate a filtered time-domain PVP signal (such as the cleaned PVP signal 508 or the conditioned PVP signal 510). The filtered time-domain PVP signal may then be transformed to the frequency domain and analyzed according to the methods discussed herein to determine one or more patient state metrics.

The filtering method 600 begins by obtaining a time-domain PVP signal from measurements associated with the pressure in the peripheral veins of the patient 102 (block 602). The time-domain PVP signal may be generated directly by the pressure sensor 112 or may be derived from sensor measurements, as discussed elsewhere herein. As described elsewhere herein, the time-domain PVP signal may be obtained by monitoring the pressure sensor 112 or by accessing a stored PVP data signal. In some embodiments, the evaluation unit 118 may monitor and record data from the transducer to generate a time domain PVP signal. The time-domain PVP signal may include one or more of each of the following: (i) an active time segment during which the pump 111 is operating (i.e., actively pumping) and (ii) an idle time segment during which the pump 111 is not operating (i.e., not actively pumping). The active time segments and the idle time segments may alternate periodically or aperiodically. While the pump 111 may be configured to operate in such a manner that both active and idle time segments are inherently present during normal use, an active time segment is a period in which the pump 111 is generating noise artifacts through active operation, whereas an idle time segment is a period in which the pump 111 is not generating significant noise artifacts through passive or inactive operation (e.g., a rest period between periodic pumping). To enable further analysis of the measured PVPs, the evaluation unit 118 may identify and filter the active time segments and the idle time segments.

Accordingly, the filtering method 600 may identify the value of the time-domain PVP signal associated with the active time segment or the idle time segment (block 604). The evaluation unit 118 may automatically identify the active time segment, the idle time segment, or both the active time segment and the idle time segment based on the value of the time-domain PVP signal. In a preferred embodiment, the evaluation unit 111 may identify the time segments based on analysis of the time-domain PVP signal alone, without referring to additional extrinsic information about the characteristics or operating state of the pump 111 (e.g., previously determined pump operating parameters or control signals controlling the operation of the pump) that is not contained in or derived from the time-domain PVP signal. Thus, the evaluation unit 118 may identify the time segments in the same manner regardless of the characteristics, configuration or settings of the pump 111 and without the need for an adjustment or further configuration of the evaluation unit 118. In various embodiments, the evaluation unit 118 may automatically identify the time segments based on the magnitude of the values of the time-domain PVP signal or based on a variation in the magnitude of the values of the time-domain PVP signal. The values may be analyzed individually or in sets containing multiple values according to one or more set metrics applied to the sets.

For individual values of the time-domain PVP signal, each of the plurality of values may be compared to one or more threshold levels to determine whether the value is associated with a time within the active time segment or the idle time segment. For example, values above an upper threshold level may be identified as being associated with active time segments, or values below a lower threshold level may be identified as being associated with idle time segments. Values may be grouped based on such comparisons to identify active time segments and idle time segments. While in some embodiments the upper threshold level and the lower threshold level may be the same, in other embodiments they may be different levels. When different, there is an uncertainty range where no value can be assigned to either the active time segment or the idle time segment. Such uncertainty values may be further analyzed based on surrounding time segments to determine whether such uncertainty values belong to an active time segment, an idle time segment, or a transition time segment. In some embodiments, outliers may be discarded or identified as part of an active or idle time segment based on values that temporally surround such outliers (i.e., before and after the outliers). To more fully remove noise artifacts, the transition time segments may be considered active time segments in some embodiments for the purpose of generating a filtered time-domain PVP signal.

For sets of values of the time-domain PVP signal, each set may be analyzed using one or more set metrics to determine whether the set is associated with an active or idle time segment. In a preferred embodiment, each set contains values of the time-domain PVP signal that are adjacent in time, thereby forming a time series of values of the PVP signal. Thus, each set is associated with a set-specific time period and includes values associated with times within that set-specific time period. The set-specific time periods of the set may cover a fixed duration or may have a variable duration, and the set-specific time periods may be overlapping or non-overlapping. The sets may include sample values from the time-domain PVP signal, or the sets may include all values of the time-domain PVP signal associated with times within a set specific time period of the corresponding set. In a particularly preferred embodiment, the set-specific time periods may be non-overlapping, but adjacent sets cover all time periods within the analysis time period for which data is available for an uninterrupted duration of the time-domain PVP signal, such that each value of the time-domain PVP signal within the analysis time period is in an exact one of the sets. Thus, an active time segment or an idle time segment may be identified as a collection of one or more sets by identifying the set as being associated with the active time segment or the idle time segment.

To identify a set as being associated with an active or idle time segment, one or more set metrics may be used to evaluate the values of the time domain PVP signals within the set. The set metric may comprise a function that determines a mean, a maximum, a minimum, a distance between the maximum and the minimum, an average change between values (or their absolute values), a variance of the set, or another metric of the values in the set. Once a set metric has been determined by evaluating the values of the set, the set metric may be compared against a set threshold level associated with the set metric to identify the set as being associated with an active time segment or an idle time segment. For example, a set may be identified as being associated with an active time segment when the set metric is above a set threshold level for the set metric and associated with an idle time segment when the set metric is below the set threshold level for the set metric.

In some embodiments, the set metric may determine a change, such as a rate of change, between values within the set. Such a rate of change may be an average rate of change, a maximum rate of change, or other measure of change between values. The aggregate measure of change or rate of change between the relevant values may be used to determine the start time or end time of an active or idle time segment by comparison against a threshold associated with the start or stop of active pumping. PVP can surge when pump 111 begins active pumping at the beginning of the active time segment and abruptly drops when pump 111 stops active pumping at the end of the active time segment. Thus, large and rapid changes in the values of the time-domain PVP signal can be used to identify the start or end of the active and idle time segments. For example, the start time of an active time segment may be identified by determining that a change or rate set metric is above a pumping start threshold, and the start time of an idle time segment may be identified by determining that a change or rate set metric is below a pumping stop threshold. Active and idle time segments may then be identified based on such start or end times.

Once the active time segment and the idle time segment are identified in the time-domain PVP signal, the evaluation unit 118 may generate a filtered time-domain PVP signal (block 606). The filtered time-domain PVP signal may be the adjusted PVP signal 510 with the estimated PVP signal 502E (as illustrated in fig. 5C) or may alternatively be the cleaned PVP signal 508 (as illustrated in fig. 5B) which simply removes the active pump PVP signal 502A. The filtered time-domain PVP signal is generated based on the time-domain PVP signal and excludes values of the time-domain PVP signal associated with the active time segment. In contrast to other methods that attempt to correct for pump noise artifacts by estimating and removing the noise artifacts themselves, the filtering method 600 estimates what the PVP signal would be if the pump 111 was not already operating.

As illustrated above in the cleaned time-domain PVP signal 508, a filtered time-domain PVP signal may be generated by removing values associated with the one or more identified periods of activity from the time-domain PVP signal. Where the time-domain PVP signal comprises a sequential time series of discrete values, the filtered time-domain PVP signal may be generated by removing those values whose times, corresponding to those values, are identified as falling within the active time segment, thereby leaving one or more sequential time series of discrete values corresponding to times falling within the idle time segment. In some embodiments, the filtered time-domain PVP signal may be further adjusted or normalized prior to further analysis. For example, the remaining values associated with the idle time segments may be pieced together to avoid having gaps (such as gap 506) in the filtered time-domain PVP signal. To do this, corresponding values within each of a plurality of idle time segments may be identified, and the idle time segments may be combined by aligning the identified corresponding values. Thus, the start of one idle time segment may be aligned with the end of the previous idle time segment such that the cycles (i.e., the patient's cardiac cycles) are aligned. This may further require removing or blending overlapping values of one or both of the idle time segments to produce an uninterrupted filtered time-domain PVP signal.

As illustrated above in the adjusted PVP signal 510, the filtered time-domain PVP signal may alternatively be generated by replacing values associated with one or more of the identified activity periods with replacement values. The replacement value is determined based on values associated with one or more idle time segments in the time-domain PVP signal. Thus, the filtered time-domain PVP signal may be generated by combining the values of the time-domain PVP signal associated with the idle time segment with the substitute values for the active time segment to produce an uninterrupted signal or time sequence of values. In some embodiments, the replacement values may be generated by estimating values for the active time segments based on a model determined by regression analysis, principal component analysis, or similar techniques. The model parameters may be estimated by ordinary least squares regression of the values associated with the idle time segments. However, in a preferred embodiment, the model may be estimated by a minimum cubic regression of the values associated with the idle time segments, which in many cases yields improved results for the PVP signal. In some embodiments, the override value may be adjusted near the boundary between the active time segment and the idle time segment in order to smooth the transition between the idle time segment value and the override value. In further embodiments, a replacement value may be estimated for both the active time segment and the idle time segment, in which case the values of both may be replaced with the estimated replacement value to generate the filtered time-domain PVP signal. Such a filtered time-domain PVP signal may be beneficial in some cases, as such a signal avoids interruptions or discontinuities in the signal at the boundary between the active time segment and the idle time segment.

Once the filtered time-domain PVP signals have been generated for one or more time periods, such as the evaluation window discussed above, the evaluation unit 118 may further analyze the data by generating frequency-domain PVP data from the one or more filtered time-domain PVP signals (block 608). In a manner similar to that discussed elsewhere herein, a time-frequency transform (such as an FFT) may be applied to the filtered time-domain PVP signal to generate frequency-domain PVP data as a representation of the PVP in the frequency domain after filtering to remove noise artifacts from operation of the pump 111. Such frequency-domain PVP data may be generated as a frequency distribution associated with one or more filtered time-domain PVP signals. By generating frequency domain PVP data using the filtered time domain PVP signal, the PVP can be analyzed for patients connected to the periodically operating pump 111, regardless of the noise artifacts generated by the pump operation. If the pump 111 is connected directly to the circulatory system of the patient,the methods described herein enable analysis at an operational rate up to a point where idle time segments become too short and rare for reliable filtering (e.g., for using SIGMA such as produced by cents international corporation)Infusion systems such infusion pumps are about 250 cc/min for most adult patients with typical heart and respiratory rates). In one embodiment, the evaluation unit 118 further normalizes the frequency domain PVP data. For example, the evaluation unit 118 may normalize the frequency domain PVP data to account for the idle time segment. The frequency domain data may then be further analyzed to determine one or more patient state metrics (block 610). Such frequency domain analysis may include the analysis of frequency peaks (F)_N) As discussed in more detail elsewhere herein. In some embodiments, this may include comparing the frequency domain PVP data to determine a change in the patient state metric.

Because a comparison of the changes in frequency and associated amplitude is particularly useful for monitoring a patient condition via a patient state metric, a discussion of such a comparison is described next. In further embodiments, similar methods of comparing metrics of the PVP signal across multiple time periods in the time domain may be performed equally to monitor patient condition. Fig. 7 illustrates an exemplary PIVA comparison methodology 700 for identifying changes in patient state based on comparison of frequency domain representations of PVP signals associated with different times. The PIVA comparison method 700 may be implemented by the evaluation unit 118 and the response unit 116 to determine and respond to changes in the patient state between time periods. For example, the evaluation unit 118 may determine and compare frequency domain representations of the PVP based on the electronic pressure signals received during the plurality of time periods to determine changes in patient metrics, such as blood pressure, blood volume, respiration, positioning or movement, or systemic vascular resistance. In particular, the evaluation unit 118 may compare the peak value (P) with the frequency distribution determined in each time segment_N) Frequency (F)_N) Correlated relative or absolute amplitude to identify changes in the state of a patientThe change in state may be used by response unit 118 to determine and implement a response action.

The example method 700 begins by obtaining a first frequency distribution associated with a first time period (block 702) and obtaining a second frequency distribution associated with a second time period (block 704). As described above, each of the first and second frequency distributions may be generated by the method 300 or the filtering method 600 as frequency domain data corresponding to a plurality of data values from the PVP data signal. The first and second time periods may correspond to first and second evaluation windows, each associated with a plurality of data values sampled or received by the evaluation unit 118, as discussed above. As discussed above, the data values for each of the first evaluation window and the second evaluation window may be stored in volatile or non-volatile memory until needed by evaluation unit 116 to generate the frequency distribution. Alternatively, the frequency distribution or information associated with the frequency distribution (e.g., frequency peaks and associated amplitudes) may be stored directly for comparison. In some embodiments, the first and second frequency profiles may be frequency domain representations of the PVP signal from the sensor 112 over a time period of fixed duration starting at times separated by a predetermined interval. For example, the frequency peaks (F) of the frequency distribution generated during patient monitoring within the first and second evaluation windows may be compared_N) To implement the method 700 on a rolling basis (i.e., periodically or when new PVP data becomes available) during real-time monitoring of the patient 102. The first and second periods of time may partially overlap, be adjacent in time or separated by an intermediate period of time.

The evaluation unit 116 may next identify one or more peaks of interest for use in determining a patient state metric (block 706). A peak of interest may be identified in one or both of the first frequency distribution and the second frequency distribution. In some cases, the peak (P) in the baseline frequency distribution generated for the patient 102 may be based on_N) To determine one or more peaks of interest, the baseline frequency distribution may be a first frequency distribution or an additional frequency distributionThe prior frequency distribution is added. The baseline frequency distribution may be determined, for example, prior to a scheduled procedure to establish a baseline for later patient status monitoring. May be based on an associated frequency (F)_N) (such as by identifying the breathing frequency (F)₀) Or heart rate frequency (F)₁) To identify the peak of interest. In some embodiments, the peak of interest may include a plurality of such peaks, such as at the first harmonic frequency (F)₂) And the second harmonic frequency (F)₃) Associated peak value (P)₂) And (P)₃). In some cases, it may not be possible to identify all peaks of interest in both frequency distributions. For example, during acute failure of the circulatory system, systemic vascular resistance may be significantly reduced and related to harmonic frequencies (F)₂、F₃、…F_N) The associated peak may not be discernable. Therefore, the harmonic frequency (F)₂、F₃、…F_N) The associated peak may be identifiable in the first frequency distribution and not in the second frequency distribution. However, the harmonic frequency (F) may be determined by comparing the first frequency distribution and the second frequency distribution₂、F₃、…F_N) The amplitude of the frequency distribution changes.

Based on the identified one or more peaks of interest, the evaluation unit 116 may further determine the patient state (or the change in the patient state) by comparing the first frequency distribution and the second frequency distribution (block 708). Determining the patient state may include pairing the same one or more frequencies (F) as between the first frequency distribution and the second frequency distribution_N) A comparison of the associated amplitudes, a comparison of values of a function of a plurality of amplitudes associated with frequencies between the first frequency distribution and the second frequency distribution (e.g., a comparison of ratios of peak amplitudes), a comparison of one or more peaks between the first frequency distribution and the second frequency distribution (P)_N) Associated frequency (F)_N) (e.g., changes in respiratory rate or heart rate) or other metrics associated with the patient's state. In some embodiments, the patient state may be determined based on the metric change exceeding a threshold level. For example, with a second frequency divisionHeart rate frequency (F) in the cloth₁) The associated amplitude is reduced to a heart rate frequency (F) in a first frequency distribution₁) Below 80% of the associated corresponding amplitude may indicate hypovolemia in the patient 102. As another example, dependent on heart rate frequency (F)₁) Whether and how the associated amplitude changes, the sum first harmonic frequency (F) between the first frequency distribution and the second frequency distribution₂) The associated amplitude and heart rate frequency (F)₁) A reduction in the ratio of the associated amplitudes beyond a predetermined threshold may be indicative of high or low blood volume. Comparisons of particular interest are discussed in more detail elsewhere herein.

With respect to patient hemodynamic state or blood volume, the harmonic frequencies (F) are involved_H) Of particular interest is the comparison of one or more of these. Because of the harmonic frequency (F)_H) Correlating the frequency distribution value with the heart rate frequency (F)₁) The associated value is more sensitive than the blood volume change, so monitoring is with harmonic frequency (F)_H) The change in the associated value may provide an earlier or clearer indication of the patient's hemodynamic status. E.g. with heart rate frequency (F) in the same patient at the same time₁) Corresponding change of the associated value, compared to the first harmonic frequency (F)₂) A sharp increase or decrease in the magnitude of the values of the associated frequency distribution (or other harmonic frequencies) may be more pronounced. Thus, harmonic frequencies (F) may be used_H) A blood volume metric is generated. Such a measure may be determined as a harmonic frequency (F)_H) Function of (2), harmonic frequency (F)_H) Ratio of frequency values of (a) to harmonic frequencies (F)_H) Associated amplitude, and harmonic frequency (F)_H) The ratio of the associated amplitudes or a change in any of these. Such changes may be measured at regular intervals in time prior to the current value, against a baseline, or against a previously determined value. In some embodiments, other correlation values (such as with respiratory rate frequency (F)) may be contrasted₀) Or heart rate frequency (F)₁) Associated frequency or amplitude) to compare with the harmonic frequency (F)_H) The associated frequency or amplitude. For example, by controlling heart rate frequency (F)₁) To one orMultiple harmonic frequencies (F)_H) And (6) carrying out normalization. Such a normalized value may be determined as a ratio of the amplitudes and may be used as a blood volume metric to assess the hemodynamic status of the patient 102. In various embodiments, a frequency (F) based at least in part on one or more harmonics may be determined and used_H) Other similar blood volume measures of frequency and amplitude values to assess the hemodynamic state of the patient 102.

Once the patient state has been determined, the response unit 116 may determine whether a response is required and cause any required response to be implemented (block 510). This may include determining a patient condition based on the patient state metric. Additionally or alternatively, the evaluation unit 118 or the response unit 116 may cause an indicator of the determined patient state to be stored or presented via the monitor 120 (block 510). If the response unit 116 determines that a response is required, the response unit 116 may further determine one or more responses suitable for addressing the identified patient state. Such a response may include generating an alarm or other alert that the patient state is abnormal, which may include information about the patient's condition. The alert or warning may be presented via monitor 120 or may be communicated to another device for presentation. The alert or warning may include a recommendation for one or more actions to be taken in response to the patient state. For example, the recommendation may include an adjustment to a fluid therapy for the patient 102, which may include a recommendation to administer one or more vasopressors or vasodilators. Such recommendations may be determined by the response unit 116 as part of the required response. In some embodiments, this may include sending an electronic communication to a user device (e.g., a workstation or mobile device used by a physician, nurse, or technician to monitor a patient's condition).

The response may similarly include controlling the fluid source 110 to regulate fluid flow to the patient 102. The fluid source 110 may be controlled to increase or decrease the rate of fluid flow to the patient 102, including starting or stopping fluid flow. In some embodiments, the response may include controlling the fluid source 110 (or a device connected to the fluid source 110) to administer one or more drugs to the patient 102. For example, the fluid source 110 may be controlled to administer one or more vasopressors or vasodilators in the fluid delivered to the peripheral vein 108 via the IV line 104 and the venous access device 106. Where the fluid source 110 includes a pump, the response may include controlling operation of the pump, such as by increasing or decreasing pump speed, flow rate, or mode of operation, and starting or stopping the pump. In some embodiments, the fluid source 110 may be controlled to administer an amount of medication to the patient 102 via a fluid. For example, the fluid source 110 may be controlled to add an amount of medication to the fluid. Additional embodiments of specific analysis and response methods utilizing the PIVA system 100 are further described in more detail elsewhere herein.

PIVA module

The PIVA system 100 can perform several signal filtering and signal processing steps (e.g., removing noise artifacts from the physiological signal, performing an FFT on the physiological signal, calculating the PIVA score via the equations previously disclosed herein as an inference of pulmonary capillary wedge pressure, and other related functions). In one embodiment, the PIVA system 100 performs these and other steps via the PIVA module 800. Although the PIVA module 800 is described with reference to the block diagram illustrated in FIG. 8, it should be appreciated that many other configurations and methods of performing the behavior associated with the PIVA module 800 may be used. For example, the order of some blocks may be changed, some blocks may be combined with other blocks, and some blocks described may be optional.

As illustrated in fig. 8, the PIVA module 800 includes a noise module 802, a signal quality index module 804, a pulse rate module 806, an FFT module 808, and a respiration rate module 810.

The PIVA module 800 receives at least one input. For example, the PIVA module 800 may receive a digital signal from an analog-to-digital converter. The digital signal may represent a patient physiological parameter, such as the patient's peripheral intravenous pressure. It should be appreciated that many other physiological parameters are contemplated, such as other invasive venous pressures, invasive arterial pressures, non-invasive venous pressures, non-invasive arterial pressures, and other similar parameters. In one example, the digital signal is derived from a medical device, such as a pressure transducer fluidly connected to a vein of a patient.

Likewise, the PIVA module 800 delivers output. For example, the PIVA module 800 can output a Signal Quality Index (SQI) related to the PIVA system 100, a patient's Respiratory Rate (RR), a patient's Pulse Rate (PR), and a patient's PIVA score.

Noise module

The PIVA module 800 may perform filtering and processing in response to receiving the digital signal. In one embodiment, the digital signal is processed via the noise module 802 to eliminate noise artifacts, such as those associated with operation of the pump. For example, the noise module 802 may perform a forward-backward slope calculation to identify segments of the digital signal where noise is present. In one embodiment, the noise module 802 performs several processing steps to eliminate noise artifacts from the signal. In one embodiment, the processing includes cascaded stack processing. This may advantageously provide real-time processing and efficient decimation for regression feature calculation, block processing, filtering, etc.

More specifically, the noise module 802 may evaluate the digital signal, identify points where the positive slope of the signal is greater than a particular threshold (e.g., a signal spike), and characterize this portion of the digital signal as the noise onset point. This can be generally characterized as slope-based burst detection. Similarly, the noise module 802 may evaluate the digital signal, identify a point at which the negative slope of the digital signal is less than a particular threshold (e.g., signal drop) and characterize this portion of the signal as the noise end point. The slope may be calculated by taking the derivative of the digital signal.

In one example, the noise module 802 implements a sliding window stack size sufficient for local parameter estimation (e.g., for real-time processing). The noise module 802 determines a slope window size on each side of a peak (e.g., a peak typically associated with signal noise) within a particular stack. For example, the slope is calculated:

ForwardSlope＝S{X[p-wdex]-X[p]}/(p-wdex)

BackwardSlope＝S{X[p]-X[p-wdex]}(p-wdex)

preferably, the spacing between the slope windows is tested for a wide range of pump rates. The noise module 802 may also calculate a point of symmetry between the forward slope and the backward slope. The symmetry point can infer peak noise localization. In one embodiment, high slope and/or high amplitude noise is detected.

Slope-based burst detection is an adaptive input signal conditioning process that provides for real-time noise cancellation. For example, the noise module 802 identifies a noise start time and a noise stop time, and removes a signal (e.g., a cascade signal) between the noise start time and the noise stop time. In other words, once a noise segment is identified (e.g., the portion of the signal between the noise start point and the noise end point), the noise module 802 may delete the segment from the digital signal (e.g., to produce a concatenated signal or a segmented signal).

Likewise, the noise module 802 may also perform image matched filtering to fill gaps in the cascaded signal, for example. More specifically, the signal range (e.g., signal noise region) between the sign-adjusted forward slope and backward slope that is greater than the threshold is replaced with a mirror image of the symmetrically segmented adjacent regions. In one embodiment, mirror matched filtering involves filling each gap from the front (e.g., from the noise end point) and from the back (e.g., from the noise start point). In various embodiments, image matched filtering involves filling gaps with prior digital signal data stored in a memory (e.g., a buffer memory). For example, the noise module 802 retrieves the buffer stack memory and fills the synthesized data from the forward and/or reverse direction of the signal. In one embodiment, the buffer and window sizes are optimized for pump rates from 25Hz to 250 Hz.

The processing performed by the noise module 802, including slope-based burst detection and subsequent image matching, advantageously eliminates noise artifacts from the signal. For example, with pump rates up to 250mL per hour, the noise module 802 has at least 0.74 seconds between pumping intervals; this is necessary to obtain an appropriate signal for patients with low pulse rates. Preferably, the end result is a cleaned-up signal with the noise artifact removed. After the noise block 802, the PIVA block 800 may perform additional processing on the cleaned signal.

Signal quality index module

In one embodiment, the cleaned signal may be processed via the signal quality index module 804 to obtain the SQI associated with the PIVA system 100. For example, the signal quality index module 804 may include an autocorrelation of the cleaned signal (e.g., waveform), which may include determining both a mean of zero crossings and a standard deviation of the zero crossings. Zero crossing analysis can be advantageously used to calculate the SQI. The PIVA module 800 may output the SQI in response to processing via the signal quality index module 804.

More specifically, determining the signal quality includes analyzing an autocorrelation of the signal. Autocorrelation may include placing the original digital signal on itself (e.g., placing the original digital signal on the cleaned signal). When there is a statistical spread in the zero crossings that is approximately the same as the zero crossing rate, the signal may be a unwanted signal. For example, when the standard deviation of zero crossings is similar to the number of zero crossing events, the signal may be useless. Calculating the signal quality:

ZCSD＝Autozerocross–zerocrossSD

signal quality sqrt (abs (ZCSD)/(autozeroscos + zerossSD))

This calculated signal quality value may be displayed as a percentage of signal quality and delivered as an SQI.

In one embodiment, a monitor in communication with the PIVA module 800 will display a specific graphical user interface if the signal quality is determined to be "low" quality. For example, the monitor may indicate "poor signal quality". Similarly, the monitor may include signal quality troubleshooting recommendations. For example, the monitor may recommend (1) checking patient status, (2) checking the IV catheter for displacement, air, and kinking, (3) checking pump rate to ensure it is below 250mL per hour, (4) checking patient movement, (5) identifying that the device may be incompatible for use with more than one infusion pump, and (6) refreshing and confirming IV catheter retraction.

Pulse rate module

In a related embodiment, the cleaned signal may be processed via the pulse rate module 806 to obtain the patient's PR. For example, the pulse rate module 806 may use a bilateral slope detection to determine the top spectral peak of the cleared signal. In an example embodiment, the two-sided slope detection is in the form of band pass filtering (e.g., high pass and/or low pass filters) implemented in hardware or software. The PIVA module 800 may output PR in response to processing via the pulse rate module 806.

More specifically, the processing includes a cascade stack processing. This may advantageously provide real-time processing and efficient decimation for regression feature calculation, block processing, filtering, etc.

In one embodiment, the pulse rate module 806 implements an autocorrelation process for periodic determinations to calculate a pulse rate (also referred to herein as heart rate or HR). For example, the pulse rate module 806 uses 8192 sample block sizes, which may also be the stack buffer size that is processed in reverse order to get the correct periodicity characteristics. The pulse rate module 806 may implement an optional overlap interval. As an example, the default interval may be a one second interval having 500 samples. The pulse rate module 806 may calculate an autocorrelation for the lag (e.g., 0 to 4000, related to a periodicity of up to 8 seconds). The pulse rate module 806 can calculate the lag of the peak associations (e.g., the lag of 17 peak associations) using the forward and backward slope detection as previously described. The pulse rate module 806 may filter the zero crossing period and the standard deviation. The pulse rate module 806 may calculate a filtered average spacing between sub-harmonics. The pulse rate module 806 may calculate an HR estimate.

In one embodiment, the pulse rate module 806 implements spectral processing (FFT) to determine HR. For example, the pulse rate module 806 uses an 8192 point block size, which 8192 point block size may preferably not include a window function. The pulse rate module 806 may determine the spectral peaks through forward and backward slope techniques. The pulse rate module 806 may use the zero order harmonic as part of the HR estimate. Peaks associated with spectral magnitudes that are independent of the repetition rate of the autocorrelation inference are identified accordingly. Spectral amplitude peak identification may be used to calculate a respiration or pulse rate (e.g., via one discrete peak) and a volume index or patient fluid status (e.g., via multiple peaks). A discussion of spectral magnitude peak identification is included in the FFT block section below.

In a related embodiment, the pulse rate module 806 implements an FFT to refine the HR previously determined via autocorrelation. In this embodiment, the initially calculated HR via autocorrelation is a partial HR estimate.

In another embodiment, the pulse rate module 806 also calculates Heart Rate Variability (HRV) and HRV variability. For example, because the pulse rate module 806 is performing peak detection over a sliding window, the pulse rate module 806 may determine how the data changes or varies, and thus determine the HRV and HRV variability.

FFT module

In a related embodiment, the cleaned signal may be processed via the FFT module 808 to obtain the PIVA score for the patient. For example, FFT module 808 may perform spectral analysis on the cleaned signal to obtain the amplitude. These FFT magnitude spectra can be used to calculate the PIVA score (as described in more detail below). The PIVA module 800 may output a PIVA score in response to processing via the FFT module 808.

More specifically, the FFT module 808 is used to identify spectral magnitude peaks, which are then used to calculate a capacity index (e.g., a plurality of peaks). In one embodiment, the processing includes cascaded stack processing. This may advantageously provide real-time processing and efficient decimation for regression feature calculation, block processing, filtering, etc.

The FFT module 808 implements spectral processing to identify spectral magnitude peaks. In one embodiment, the identification of individual amplitude peaks comprises: the amplitude peak of the fourier transform is found with the maximum found forward-backward slope change aided by the guidance of the autocorrelation pulse rate.

c₀、c₁、c₂、c₃、g₀、g₁、g₂、g₃、h₀、h₁、h₂、h₃、i₀、i₁、i₂and i3 are each a constant. mag^f1、mag^f2And mag^f3Each representing a respective frequency (e.g., F)₁、F₂、F₃) An individual amplitude of each of the. These amplitudes are also commonly referred to herein as peaks in frequency. For example, mag^f1May also be referred to herein as peak and heart rate frequency F₁Associated P₁. Similarly, for example, mag^f2May also be referred to herein as a harmonic of the first harmonic frequency F₂Associated peak value P₂. Similarly, for example, mag^f3May also be referred to herein as harmonic of the second harmonic frequency F₃Associated peak value P₃. For example, and referring to FIG. 4B, it is referred to as mag in the PIVA equation^f1P of₁Is the heart rate frequency (F)₁) Is called mag in the PIVA equation^f2P of₂Is the first harmonic frequency (F)₂) And is referred to as mag in the PIVA equation^f3P of₃Is the first harmonic frequency (F)₃) Of the amplitude of (c).

Additional ways to determine the relationship between the PIVA score and the patient's pulmonary capillary wedge pressure include evolutionary algorithms to fit the data to optimize low complexity and low error solutions and neural network mapping the data with training and validation sets using nodes of hyperbolic tangent functions to create non-linear relationships between values.

In a related embodiment, FFT module 808 performs an algorithmic method of calculating the capacity index. For example, the FFT module 808 performs an initial least squares method to analyze individual amplitudes (e.g., F)₁、F₂、F₃Etc.) and then calculate a best fit for the capacity index. Alternatively, the best fit for the volume index may be characterized as the best fit for the pulmonary capillary wedge pressure. In response to generating the best fit, the FFT module 808 may use this best fit for subsequent iterations to calculate the capacity index. In this example, subsequent iterations may allow additional calculations of the PIVA score to be implemented.

Respiration rate module

In one embodiment, the cleared signal may also be processed via the respiration rate module 810 to obtain the RR of the patient. For example, the respiration rate module 810 may filter the cleaned signal through a high pass filter. The respiration rate module may further perform a regression dispersion analysis (e.g., sin () + cos ()) and calculation of the associated ArcTan (y/x) to determine the RR. The PIVA module 800 may output an RR in response to processing via the respiration rate module 810.

Determining the respiration rate may include using a digital linear FM discriminator based on differential phase angle filtering. As described above, before this determination, the pulse rate is calculated. The pulse rate data is then replicated. The respiration rate module 810 applies a digital high pass filter to the signal. For example, a high pass filter isolates the breathing rate frequency range and allows fitting of the data to extract the breathing rate.

More specifically, the input signal is high pass filtered for maximum volatility detection. The respiration rate module 810 performs orthogonal regression filtering:

cosine (2 × PI n × k) and

Sin(2*PI*n*k)

and then calculates a filtered ArcTan () of the filtered orthogonal term. The respiration rate module 810 calculates the derivative of the filtered ArcTan () angle. In one embodiment, the respiration rate module 810 further lightly filters the derivative of the filtered ArcTan () angle. The respiration rate module 810 may then estimate the dominant baseband frequency. Multiplying the estimate by 60 provides the respiration rate on a per minute basis.

In various embodiments, the RR of the patient is determined directly via the FFT signal. For example, as stated previously and with reference to FIG. 4B, under normal conditions, there is the lowest frequency (F)₀) Peak value (P) of₀) Corresponding to the breathing rate of the patient 102. Likewise, having the next lowest frequency (F)₁) Peak value (P) of₁) Corresponding to the heart rate of the patient 102. Thus, the respective peak values P can be directly passed through₀And P₁Is easy to determine the RR (and HR) of the patient.

PIVA system

Fig. 9 illustrates a block diagram of an exemplary PIVA system 900 including the PIVA module 800 previously described herein. In addition to the PIVA module 800, the PIVA system 900 can also include a processor 902 and memory 904 running on the PIVA module 800. For example, the PIVA module 800 may include one or more physical processors 902 communicatively coupled to one or more memory devices 904.

A physical processor, such as processor 902, refers to a device capable of executing instructions encoding arithmetic, logical, and/or I/O operations. In one illustrative example, a processor may follow a von neumann architecture model and may include an Arithmetic Logic Unit (ALU), a control unit, and a plurality of registers. In one example, the processor may be a single core processor that is typically capable of executing one instruction (or a single pipeline that processes instructions) at a time or a multi-core processor that can execute multiple instructions simultaneously. In another example, the processor may be implemented as a single integrated circuit, two or more integrated circuits, or may be a component of a multi-chip module (e.g., where separate microprocessor dies are included in a single integrated circuit package and thus share a single socket). The processor may also be referred to as a Central Processing Unit (CPU). A memory device, such as memory device 904, refers to volatile or non-volatile memory devices, such as RAM, ROM, EEPROM, or any other device capable of storing data. Local connections, including connections between the processor 902 and the memory device 904, may be provided by one or more local buses of a suitable architecture, such as Peripheral Component Interconnect (PCI).

Likewise, the PIVA system 900 can include sensors 906 and monitors 908. For example, the PIVA module 800 may communicate with each of the sensors 906 and the monitor 908. The communication may be wired and/or wireless (e.g., WiFi, bluetooth, and other related wireless protocols). In one example, the sensor 906 is the pressure sensor 112 described in more detail above. In one example, the monitor 908 is the monitor 120 described in more detail above. In one embodiment, the PIVA module 800 is physically located within the monitor 908.

Likewise, the PIVA system 900 may include a database 910 and a cloud 912. For example, the PIVA module 800 may communicate with each of the database 910 and the cloud 912. The communication may be wired and/or wireless (e.g., WiFi, bluetooth, and other related wireless protocols). In one example, database 910 includes electronic medical records stored on a hospital network. In one example, the cloud 912 includes a remote storage location that can be used to store physiological data and/or device information (e.g., PIVA module 800 performance statistics, software updates, and other relevant information).

In one embodiment, the PIVA system 900 displays the updated capacity index via the monitor 908 every 60 seconds. Preferably, the PIVA fraction used to calculate the capacity index is consistent with a pulmonary capillary wedge pressure of ± 8mmHg at the limit of consistency of the 95% confidence interval.

In one embodiment, the PIVA system 900 displays the updated pulse rate via the monitor 908 every 10 seconds. Preferably, the pulse rate has a coherence with a heart rate of ± 10 beats per minute at the limit of coherence of the 95% confidence interval.

In one embodiment, the PIVA system 900 displays the updated breathing rate via the monitor 908 every 10 seconds. Preferably, the respiration rate is consistent with a respiration rate of ± 5 breaths per minute at the limit of consistency of the 95% confidence interval.

In one embodiment, the PIVA system 900 operates in conjunction with an external medical device. For example, the PIVA system 900 operates in conjunction with an infusion pump operating at a rate of 0 to 250mL per hour. In a related embodiment, the PIVA system 900 utilizes noise cancellation (e.g., via the noise module 802) to remove the pump signal from the detected waveform (e.g., digital signal).

In one embodiment, the PIVA system 900 displays a volume index (e.g., PIVA score), a pulse rate, and a respiration rate when signal quality is sufficient. For example, signal quality may be sufficient when the signal quality index indicates that the signal quality is sufficient. The PIVA system 900 may indicate that the signal has a "low" quality if the signal quality is insufficient and/or stop displaying physiological values (e.g., PR, RR, PIVA score, and other relevant physiological values) as long as the signal quality is still insufficient.

The PIVA system 900 may include other additional features. In one embodiment, the PIVA system 900 includes a power supply. The power supply may be wired to an external source and/or may have an internal power supply (e.g., a lithium ion battery). In one embodiment, the PIVA system 900 includes one or more speakers (e.g., a main speaker and a back-up speaker). The speaker may be configured to emit an alarm sound if necessary.

Fig. 10 illustrates another example of signal processing via process 1000. In various embodiments, any of the PIVA system 100, the PIVA system 900, and the master controller 1009 (as detailed below) may perform the process 1000. In one embodiment, process 1000 may be implemented in conjunction with process 800. In different embodiments, the example 1000 is a separate process different from the process 800. Although the process 1000 is described with reference to the block diagram illustrated in fig. 10, it should be appreciated that many other configurations and methods of performing the actions associated with the process 1000 may be used. For example, the order of some blocks may be changed, some blocks may be combined with other blocks, and some blocks described may be optional.

As illustrated in fig. 10, the process 1000 may include several separate functions, including an interference cancellation logic function 1002, a frequency magnitude detection function 1004, a pulse rate detection function 1006, and a respiration rate detection function 1008. Each of these functions may be performed by the main controller 1009 (e.g., a processor) or operated in conjunction with the main controller 1009 (e.g., a processor).

The interference cancellation logic function 1002 may include a sensor input (block 1010). For example, a sensor input (e.g., a pressure transducer signal) at a particular frequency (e.g., 500Hz) may be received as the sensor input. The sensor input may be an analog and/or digital signal. The 149 point FIR filter output (e.g., a low pass filter) may be added to the slope array (block 1012). For example, a filter may be added to a slope array representing the digital signal received as the sensor input. In an example embodiment, the interference cancellation logic function 1002 includes a low pass convolution filter to further improve the signal. Forward and backward slope detection may be performed (block 1014). The interference cancellation logic function may determine whether a forward/backward slope is present (block 1016). If there is a forward/backward slope, the detected data points in the slope detection array are removed (block 1018), missing values are filled in from the points before removal (block 1022), and a cubic fit is applied to remove discontinuities (block 1024). Alternatively, if there is no forward/backward slope, the input data is used as a value (block 1020).

The current point (e.g., the current point of the signal) is compared to the histogram and rejected if the current point is outside of confidence limits (block 1026). The data is "smoothed" to a cubic fit equation (block 1028). For example, a low pass filter (e.g., 16Hz) may be implemented. The output of the filtering step is added to the autocorrelation array and the samples are shifted by one (block 1030). In one example, the process repeats with 149 point FIR filter output added to the slope array (block 1012). In a different example, process 1000 proceeds to the next function.

The frequency magnitude detection function 1004 includes determining whether the sample count is greater than an FFT trigger value (block 1032).

If the sample count is greater than the FFT trigger value (block 1032), the FFT for the nearest 8192 point is performed (block 1034). The magnitude of the FFT output is calculated (block 1036). The largest peak is identified to estimate the pulse rate (block 1038). A pulse rate is calculated (block 1040). For a first frequency (e.g. F)₁) The harmonic of (a) is subjected to a peak search (block 1042). A spectral magnitude calculation is performed based on the pulse rate estimate (block 1044). A PIVA capacity index (e.g., a PIVA score) is calculated (block 1046) and an FFT trigger value is updated (block 1048). The PIVA capacity index may be sent to the master controller 1009. Because the PIVA capability index is calculated, the frequency magnitude detection function 1004 may also be generally characterized as a PIVA fractional function.

If the sample count is not greater than the FFT trigger value (block 1032), process 1000 proceeds to the next function. Likewise, in response to performing a spectral magnitude calculation based on the pulse rate estimate (block 1044), process 1000 may proceed to a next function.

The pulse rate detection function 1006 includes determining whether the sample count is greater than an autocorrelation trigger value (block 1050).

If the sample count is greater than the autocorrelation trigger value (block 1050), the inverse FFT magnitude for the nearest 8192 point is performed (block 1052). In one example, the inverse FFT magnitude provides a time domain signal (e.g., for autocorrelation). The true output is scaled to the square root of the magnitude (block 1054). The minimum, maximum, and mean values of the autocorrelation are identified (block 1056). A cubic fit of 4000 points is performed (block 1058). The forward/backward slope calculations for the minimum and maximum slope pairs are performed (block 1060). The pulse rate is calculated (block 1062). In one example, the calculated pulse rate is equal to the number of slope pairs. The Signal Quality Index (SQI) is evaluated to determine if the SQI is greater than a particular threshold (e.g., SQI > 70). If the SQI is greater than the threshold (block 1064), a weighted average of the pulse rates is calculated (block 1066) and the autocorrelation trigger values are updated (block 1068). The calculated pulse rate may be transmitted to the main controller 1009. In one example, calculating a weighted average of the pulse rates (block 1066) includes receiving an input of the pulse rates calculated by the frequency magnitude detection function 1004 (block 1040).

If the sample count is not greater than the autocorrelation trigger value (block 1050), the process 1000 continues with the next function.

The breath rate detection function 1008 includes determining whether the sample count is greater than a breath rate trigger value (block 1070). If the sample count is not greater than the breath rate trigger value, the envelope of the zero crossing is calculated (block 1072). For example, the envelope of zero crossings is scaled and normalized to the log of the autocorrelation₁₀. The envelope is compared to a template of breathing rate loaded into RAM (block 1074). The signal quality is evaluated to determine if the SQI is greater than a certain threshold (e.g., SQI > 70). If the SQI is greater than the threshold (block 1076), the respiration rate is calculated (block 1078) and the respiration rate trigger value is updated (block 1080). The calculated breathing rate may be sent to the main controller 1009.

The SQI is calculated using zero crossing statistics derived from the autocorrelation (e.g., the number of zero crossing events and the standard deviation of the zero crossing events) (block 1082). For example, calculating the SQI (e.g., at block 1064 or at block 1076) may consider the pulse rate relative to the number of slope pairs calculated by the pulse rate detection function 1006 (block 1062). A rolling average of the missing SQI over time is also calculated (block 1084). If the SQI rolling average trigger is reached (block 1086), a SQI error is sent to the master controller 1009. Likewise, if the SQI rolling average trigger is not reached (block 1086), the SQI is sent to the master controller 1009 without error.

Assessment of patient gait, epilepsy, activity and related biometrics

As previously identified with respect to fig. 1E, the example PIVA system 100 can further include one or more additional sensors 150. The one or more additional sensors 150 may be capable of being used, for example, to calculate other patient variables (e.g., in addition to the PIVA score).

More particularly, in some embodiments, the patient state metric may be directed to aspects of the patient's condition, such as patient body positioning or movement. Thus, the PVP signal may instead be used on an ongoing basis to monitor information previously monitored non-systematically by a nurse or physician's observations. For example, such monitoring may include determining one or more patient state metrics associated with the positioning or movement of the patient. In addition to the patient localization metric or the patient movement metric, the analysis of the PVP signal may also be used to generate a plurality of patient localization metrics, patient gait metrics, patient lameness metrics, patient fall metrics, patient epilepsy metrics, other patient movement metrics, patient blood volume metrics, patient vascular response metrics, patient respiration metrics, or other similar metrics associated with the patient condition described herein. In some embodiments, the PVP signal may be analyzed to generate a primary patient metric, such as pulse rate, pulse pressure, respiration rate, or depth of respiration. Thus, in some embodiments, the plurality of patient state metrics may be continuously monitored based only on PVP-related measurements, without the need for additional sensors or other types of measurements. By monitoring various patient state metrics using PIVA analysis of the PVP signal, the methods and systems described herein avoid the complexity, redundancy and incompatibility of existing systems while enabling metric-based monitoring of additional patient conditions previously monitored by human observation alone.

In still further embodiments, time domain analysis may additionally or alternatively be performed to evaluate the PVP signal. The PVP signal generated by the pressure sensor 112 may be analyzed in the time domain or in both the time and frequency domains to determine patient status or to generate patient status metrics, such as those discussed above. This may include evaluating changes in the pressure signal due to fluid movement within the IV line caused by patient movement impacting the pressure transducer in a regular pattern to determine patient movement or gait. For example, the shock associated with patient gait may create a water hammer within the IV tubing 104 that may be recognized by the evaluation unit 118 as a pattern of high and low pressure observations in the PVP signal. As another example, a measure of signal variability or variance may be generated to identify epilepsy, where the variance in the measured PVP signal will increase dramatically due to pressure changes from patient movement. Thus, analysis of the PVP signal from the sensor 112 may include identification of cyclic or acyclic patterns that may be analyzed in the time domain (e.g., by pattern recognition or identification of sudden pressure changes) or in frequency (e.g., by analysis of frequencies or amplitudes associated with local peaks in the frequency domain representation of the PVP signal).

Fig. 11 illustrates a flow diagram of an exemplary patient monitoring method 1100 that uses patient PVP to determine and respond to a patient state metric. The exemplary patient monitoring method 1100 obtains PVP data for the patient 102 and analyzes the data using the PIVA system 100. The example method 1100 may be performed by one or more software or hardware modules of the analysis component 114 using the electronic pressure signal from the pressure sensor 112, which may include generating the electronic pressure signal by the pressure sensor 112 in some embodiments. Likewise, the example method 1100 may include determining and implementing a response to one or more patient metrics, which may include presenting an alert or controlling a medical device to treat a patient condition (e.g., controlling operation of a pump or other fluid source connected to the circulatory system of the patient).

The exemplary method 1100 begins by monitoring the PVP signal of the patient 102 (block 1102). This may include generating a time-domain PVP signal by measuring a physical phenomenon associated with the PVP of the patient 102 via the IV tubing 104 using a PVP sensor, such as the pressure sensor 112 or similar sensor. Alternatively, this may include receiving or accessing a continuous or discrete time domain PVP signal from a PVP sensor or storage medium, which may include raw measurement data from the sensor or data derived from the raw measurement data. Regardless of how obtained, the PVP signal may be monitored until time-domain PVP data is obtained for a sufficient sampling period (e.g., evaluation window) to enable transformation and analysis, as discussed elsewhere herein.

Once the PVP signal has been obtained, the analysis component 114 may generate a frequency distribution by transforming the time domain PVP signal to the frequency domain (block 1104). This may include applying a Fast Fourier Transform (FFT) or other transform technique to the time-domain PVP signal, as discussed elsewhere herein. As discussed elsewhere herein, the frequency distribution may be represented in any convenient form, including an array or matrix storing data of associated frequencies and amplitudes. In some embodiments, this may include determining a plurality of frequency distributions from overlapping or non-overlapping portions of the PVP signal (e.g., the first half of the sampling period and the second half of the sampling period). Such multiple frequency distributions may then be analyzed to determine one or more patient state metrics, which may then be compared to determine a change in the patient state or condition.

It is then possible to estimate the peak (P) in the frequency distribution_N) One or more frequencies (F)_N) Or associated amplitudes, to determine at least one patient state metric (block 1106). In some embodiments, this may include identifying one or more frequencies (F) associated with local maxima of the frequency-domain PVP signal represented by the frequency distribution_N) As in the textAs discussed herein. Alternatively, the frequency distribution may comprise information indicative of such frequencies and amplitudes. Based on the identified frequency (F)_N) One or more patient metrics associated with the positioning or movement of the patient 102 may be determined by analyzing the frequencies or associated amplitudes. In some embodiments, the patient state metric may be determined based on changes in frequency or amplitude. Such a change may be determined by comparison against a previously measured frequency profile (e.g., for PVP measured during an immediately preceding sampling period) or against a baseline frequency profile (e.g., for PVP measured while the condition of the patient is known, such as immediately prior to surgery). Such a baseline frequency profile may include information about one or more baseline frequencies and associated baseline amplitudes associated with known patient conditions. The one or more patient state metrics may include a patient positioning metric, a patient movement metric, or a primary patient metric.

The primary patient metrics provide basic information about the patient 102 and may be used directly or indirectly to monitor the patient's condition. Thus, the primary patient metrics may include information about the patient's cycle and respiratory state, such as pulse rate, pulse pressure, respiratory rate, or depth of breath. By simply identifying the breathing frequency (F)₀) Or heart rate frequency (F)₁) To determine the respiration rate or pulse rate. Heart rate frequency (F)₁) Harmonic frequency (F) of₂、F₃、…F_N) Can be used to identify or confirm heart rate frequency (F)₁) Said heart rate frequency (F)₁) Can be further used for identifying the respiratory frequency (F)₀). With respiratory frequency (F)₀) Or heart rate frequency (F)₁) The associated amplitude may be used to determine depth of breath or pulse pressure. In some embodiments, the heart rate frequency (F) may be determined by comparing the heart rate frequency (F) with a predetermined threshold value₁) Associated peak value (P)₁) Is converted to a time domain signal and its amplitude is determined to determine the pulse pressure. Similarly, the peak value (P) can be obtained by matching the peak values₀) Is converted into a signal in the frequency domain, its amplitude is determined and the depth of respiration is calculated based on the amplitude to be based on the respiration frequency (F)₀) CorrelationDetermines the depth of breath. For example, the depth of breath may be calculated using a statistical model determined from PVP measurements and depth of breath measurements (or estimates) during the baseline period. Other similar primary patient metrics may be similarly determined from the frequency distribution.

Patient positioning metrics provide information about the posture or relative positioning of various parts of the patient's body. Thus, the patient positioning metric may indicate whether the patient 102 is in an upright or recumbent position. May be based on one or more frequencies (F)_N) To determine such relative positioning information as to whether the patient is standing, sitting or lying down. For example, the patient location metric may indicate a peak (P) that is the same in relation to a known patient location (e.g., sitting)₁) Compared to the heart rate frequency (F) of a previously measured amplitude₁) Associated peak value (P)₁) The absolute amplitude of the signal varies. Thus, an increased magnitude may indicate an increase in pressure in the patient's peripheral venous system, which may indicate that the patient 102 is lying down. In some embodiments, when the IV tubing 104 is connected and subsequently used in the determination of the patient positioning metric, information regarding the position of the venous access device 106 (e.g., in the patient's hand, arm, or leg) may be recorded. In further embodiments, a ratio or combination of amplitudes may be used, such as with heart rate frequency (F)₁) Associated peak value (P)₁) Sum of amplitudes and harmonic frequencies (F) thereof₂、F₃、…F_N) Associated peak value (P)₂、P₃、…P_N) Of the ratio of one or more of (a).

The patient movement metric provides information about the occurrence of patient movement, the type of patient movement, or the patient's movement-based condition. Such patient movement metrics may provide information about sudden patient movement, muscle spasm, patient gait, lameness, stability, fall, or epilepsy. The patient gait metric may be determined based on a frequency distribution associated with the PVP of the ambulatory patient. A gait frequency (F) associated with the gait of the patient while walking can be identified from the frequency distribution_G). In some embodiments, this may include first identifying the callSuction frequency (F)₀) Or heart rate frequency (F)₁) Based on a frequency lower than the heart rate frequency (F) in the frequency distribution₁) Rather than respiratory frequency (F)₀) Peak value (P) of_G) To identify the gait frequency (F)_G). In further embodiments, first of all, it is possible to pass, for example, the amplitude or the harmonic frequency (F)₂、F₃、…F_N) To identify the heart rate frequency (F)₁). Identifying gait frequency (F)_G) May also include comparing to the breathing frequency (F)₀) And gait frequency (F)_G) Relative amplitude correlated to the gait frequency (F)_G) Identified as being associated with a lower amplitude. In some embodiments, the breathing rate (F) may be based on a previous time period (e.g., a previous sampling period or evaluation window)₀) To identify the breathing frequency (F)₀). The gait frequency (F) can be further evaluated_G) To determine information about the gait of the patient, such as rate, regularity, lameness or stability.

In some embodiments, it may also be based on corresponding peaks (P) in the frequency distribution_G2) To identify a secondary gait frequency (F)_G2). Such a secondary gait frequency (F) may be determined when determining a patient gait metric or a patient lameness metric alone_G2). The secondary gait frequency (F) can be adjusted_G2) Is identified as gait frequency (F)_G) Fraction or multiple of (a). Alternatively, the secondary gait frequency (F) may be_G2) Identified as spanning a plurality of frequency distributions and gait frequencies (F) associated with a time series of sampling periods_G) Beginning and ending at the same time. As another alternative, the secondary gait frequency (F) may be_G2) Is identified as starting and in the frequency distribution with the respiratory frequency (F)₀) Heart rate frequency (F)₁) Harmonic frequency (F)₂、F₃、…F_N) Or gait frequency (F)_G) Associated corresponding peak (P)_G2) And (4) associating. In some related embodiments, a minimum threshold amplitude may be used to ensure a secondary gait frequency (F)_G2) Frequency of gait (F)_G) Of sufficient amplitude to eliminate from consideration the noise or other phenomena that may be caused by the patient wanderingA small frequency peak. Regardless of the identification, the secondary gait frequency (F) can be evaluated_G2) To determine the consistency of the gait of the patient, including whether the patient lames while walking. The frequency (F) of the secondary gait can be evaluated_G2) An associated regularity (i.e., a fixed nature of frequency) or amplitude to determine a patient gait consistency metric (which may be part of the patient gait metric) that indicates whether the patient's gait is stable, unstable, normal, or abnormal (i.e., indicates lameness). E.g. minor gait frequency (F)_G2) At gait frequency (F)_G) Consistent setting at integer fractions or multiples of (F) may indicate lameness, however, a secondary gait frequency (F)_G2) A changing frequency shift value over time may indicate instability. Similarly, with respect to gait frequency (F)_G) Associated amplitude with a secondary gait frequency (F)_G2) An associated greater magnitude may indicate a more pronounced limp home.

Although only one secondary gait frequency (F) is discussed above_G2) However, it should be understood that multiple secondary gait frequencies (F) can be identified and evaluated_G2) To determine the consistency of the gait of the patient. Similarly, in some embodiments, the gait frequency (F) may be compared against a plurality of amplitudes associated with other frequencies within a frequency distribution_G) The associated amplitude to determine the consistency of the gait of the patient. Such a comparison may be made against a measure of the total or average amplitude of the range of frequencies across the frequency distribution. For example, the gait frequency (F) may be compared_G) The ratio of the associated amplitude to the median amplitude of the frequency distribution is calculated as a patient gait metric indicative of the stability of the ambulatory patient. A higher ratio indicates a stable gait when the patient is walking, whereas a lower ratio indicates instability, as smaller changes in gait result in a relatively larger amplitude at other frequencies. Thus, frequencies not associated with the peak of the frequency distribution (i.e. local maxima) may even be evaluated when generating some patient state metrics, in particular metrics associated with stability or instability. In some embodiments, the frequency may be based on a plurality of secondary gait frequencies (F)_G2) Or other frequency-dependent amplitude (packet)Including the average (e.g., median) across the range of frequencies within the frequency distribution) to determine an individual patient stability metric.

The patient movement metric may further comprise a patient fall metric indicating that the patient has fallen. In some embodiments, the patient fall metric may be determined as a binary metric indicating the presence or absence of a fall. Alternatively, the patient fall metric may be determined as the probability of falling based on the frequency distribution. In the time domain, a fall will appear to be a sudden peak in the measured pressure, with the PVP rising rapidly as the shock of the impact propagates through the circulatory system and then returning rapidly to approximately the previous level. In the frequency domain, such spikes or pulses in the time domain signal may be identified by the characteristic pattern of peaks and troughs. For example, a spike may be considered as approximating a square pulse, and it is well known to narrow the characteristic frequency distribution of the amplitude peak symmetrically around frequency zero (0 Hz). Thus, PVP pulses associated with impacts when the patient falls can be identified by identifying patterns within the frequency distribution associated with short duration pulses in the time domain PVP signal. In some embodiments, the frequencies of interest (such as respiratory frequency (F)) in the frequency distribution may be identified and removed₀) Heart rate frequency (F)₁) Its harmonic frequency (F)₂、F₃、...F_N) Or gait frequency (F)_G) ) the associated peak to identify the pattern. In an alternative embodiment, the time domain spike may be identified as a large magnitude of the peak associated with a low frequency in the frequency distribution. Because the transient pressure pulses from a fall will be large relative to other effects on the time domain PVP signal, the amplitude associated with the main peak of the frequency distribution generated thereby will also be large. Thus, falls may be detected based on such amplitudes in some cases.

In further embodiments, spikes associated with falls may be further identified in the time-domain PVP signal, which may be advantageous in identifying the occurrence of a fall and identifying the time of the fall. Once the time of a fall is identified, the sampling period comprising the fall can be divided into a pre-fall portion and a post-fall portion for further evaluation. In some cases, the pre-fall and post-fall portions of the sampling period may be augmented by adding earlier and later values of the time-domain PVP signal, respectively, to ensure that sufficient time-domain PVP data is used for the evaluation of each portion of the original sampling period. The pre-fall part and the post-fall part can be transformed separately to generate pre-fall and post-fall frequency distributions. The frequency shift or amplitude change of the peak of interest can then be evaluated to determine the severity of the fall, which can be included in the patient fall metric. For example, the heart rate frequency (F) immediately following a fall may be determined₁) Is calculated as a measure for assessing the severity of the fall. Other similar variations in frequency or associated amplitude may likewise be determined in various embodiments.

The patient movement metric may further include a patient epilepsy metric indicative of the occurrence of epilepsy. Patient epileptic metrics may include heart rate frequency (F)₁) A ratio of the associated amplitude relative to an amplitude associated with one or more other frequencies within the frequency distribution. For example, with heart rate frequency (F)₁) The ratio of the associated amplitude to the average amplitude of the frequencies over the range of the frequency distribution (e.g., from 0Hz to 5Hz) may be used as a patient seizure metric to indicate how well-defined the heart rate is relative to the other components of the time-domain PVP signal. Heart rate frequency (F), although other factors may affect it₁) The ratio between the amplitude of (d) and the mean amplitude will be smaller for epileptic than for healthy patients. During epilepsy, the movement of the patient's body generates substantial noise in the PVP signal, resulting in a general increase in amplitude associated with frequency across the frequency distribution. If severe enough, the heart rate frequency may not be discernable from the surrounding noise (F)₁). In further embodiments, the patient epileptic metric may be determined based on an absolute level of average amplitude of the frequency distribution or an average (e.g., median) amplitude of samples from a plurality of frequencies (e.g., ten or twenty frequencies). In a related embodiment, a patient epileptic metric may be determined based on a comparison of average amplitudes between frequency distributions associated with different sampling periods, such thatA sharp increase in the mean amplitude may be indicative of epilepsy.

In some embodiments, the analysis component 114 can evaluate the peak (P) in the frequency distribution, such as by evaluating_N) Frequency (F)_N) Or associated magnitudes to monitor multiple patient state metrics simultaneously. Such a patient state metric may be determined using the same frequency distribution for the same sampling period. When a comparison between sampling periods is used to generate a patient state metric, the same plurality of frequency distributions associated with the same sampling period may be used. The plurality of patient state metrics may include metrics from one or more of the group of primary patient metrics, patient positioning metrics, or patient movement metrics discussed above, among other metrics. For example, a patient seizure metric and another patient movement metric (e.g., a patient gait metric or a patient fall metric) may be monitored simultaneously according to the same frequency distribution. As another example, a primary patient metric (e.g., pulse rate, pulse pressure, respiration rate, or depth of respiration) and a patient location metric or a patient movement metric may be monitored simultaneously according to the same frequency distribution. As yet another example, a fall or epilepsy may be identified by identifying a combination of abnormal gait metrics (e.g., variable gait frequency, minor gait frequency, or water hammer effect) and patient stress indicators (e.g., increased heart rate or respiration rate). By using (explicitly or implicitly) the frequency and amplitude information associated with the observed PVP signal, any or all of the foregoing patient state metrics may be monitored without the use of additional sensors (e.g., pressure sensor 112) other than the PVP sensor.

While the foregoing description has presented the analysis as being performed using frequency domain PVP data, other embodiments may additionally or alternatively include other types of analysis to generate patient state metrics, including any one or combination of the primary patient metrics, patient positioning metrics, and patient movement metrics discussed above. For example, a patient movement metric may be determined by analyzing the PVP signal in the time domain to identify the patient's body movement or gait by evaluating the pressure changes due to fluid movement (e.g., water hammer effect) within the IV tube caused by the patient's arm movement within the gait activity impacting the pressure sensor 112 in a regular pattern. As another example, a patient fall metric may be determined by identifying a pressure spike in the time-domain PVP signal from the pressure sensor 112 that exceeds a threshold amplitude.

Based on the one or more patient state metrics, the analysis component 114 can determine a response to the patient condition (block 1108) and implement the determined response (block 1110). For example, response unit 116 may determine whether a response is required and cause any required response to be implemented. This may include determining one or more patient conditions by evaluating one or more patient state metrics. Patient conditions may include positioning (e.g., sitting or standing), instability, lameness, fall, epilepsy, or other similar conditions. The patient condition may include a location condition, a mobility condition, or a primary condition. For example, the positioning situation may comprise lying, sitting or standing, whereas the movement situation may comprise walking, unstable walking, lameness, falling or grasping. The primary conditions may include shallow breathing, hyperventilation, non-breathing, irregular breathing, normal heartbeat, slow heartbeat, fast heartbeat, or irregular heartbeat. Determining each of the patient conditions may include evaluating one or more patient state metrics. For example, determining that the patient is walking unsteadily may include evaluating a patient gait metric to determine whether the patient is walking, and then evaluating a separate stability metric to determine that the patient is unstable while walking. Some conditions may be determined based on a combination of such metrics. For example, determining that the patient is in a normal condition may require that all monitored patient state metrics be within an acceptable range.

Whether the patient condition is determined based on the patient state metric or implied from the value of the patient state metric, one or more responses related to one or more patient conditions may be determined based on the patient state metric. While some conditions may require an active response, other conditions may simply require constant monitoring (or no response). For example, when all patient state metrics are determined for the patient 102, the analysis component 114 can determine that an appropriate response to a normal patient condition will continue to be monitored. In such cases, a response may be achieved by generating or obtaining additional sensor data about the PVP and performing further analysis on the additional data according to the methods described herein. The active response may include presenting an alarm or controlling the operation of the medical device. An alert may be generated based on the patient state metric or the condition determined thereby, which may include information about the condition or remedial action to be taken. For example, the alarm may indicate that the patient is walking erratically. A visual, audible, or tactile alert or warning may be presented (e.g., via monitor 120) to the appropriate personnel based on the alert, which may include displaying a message indicating the type of condition or recommended course of action. Operation of the medical device in response to the patient condition may include controlling the fluid source 110 to regulate the flow of fluid to the patient 102. This may include adjusting the flow rate, starting or stopping fluid flow, adding one or more drugs to the fluid, or similar control actions, as discussed further below. In some embodiments, the analysis component 114 may directly control the implementation of the response by controlling the fluid source 110 or the monitor 120. Alternatively, the analysis component 114 can transmit control information to other devices to cause those devices to present alerts or control the operation of the medical device.

In some embodiments, the patient condition or response to the patient condition may be determined based in part on additional sensor data from one or more additional sensors 150. For example, the pressure sensor may generate additional sensor data indicating whether the patient is in bed, which may be combined with patient movement metrics to determine whether the patient is at risk of falling. If the additional sensor data indicates that the patient is lying in bed, the response to continued monitoring may not need to be exceeded regardless of the instability indicated by the patient stability metric. However, if the additional sensor data instead indicates that the patient is not lying in bed, an alarm may be generated to alert appropriate personnel that the patient is in danger of falling. Some embodiments may not include additional sensors 150 or may not use sensor data from additional sensors 150 to determine a patient condition or response to a patient condition. In such embodiments, the response may be determined using patient state metrics derived solely from measurements of the PVP via the pressure sensor 112.

As used in this specification, including the claims, the term "and/or" is an inclusive or exclusive conjunction. Thus, the term "and/or" means that two or more things are present in a group or that a selection can be made from a set of alternatives.

The many features and advantages of the disclosure are apparent from the written description and, thus, it is intended by the appended claims to cover all such features and advantages of the disclosure. Further, since numerous modifications and changes will readily occur to those skilled in the art, the disclosure is not limited to the exact construction and operation as illustrated and described. Therefore, the described embodiments should be taken as illustrative and not restrictive, and the disclosure should not be limited to the details given herein but should be defined by the following claims and their full scope of equivalents, whether foreseeable or unforeseeable now or in the future.

The claims (modification according to treaty clause 19)

1. A system for monitoring a patient while the patient's circulatory system is connected to a pump using measurements associated with Peripheral Venous Pressure (PVP) within peripheral veins of the patient's circulatory system, comprising:

a PVP sensor comprising a transducer disposed adjacent to or connected to an Intravenous (IV) tube in fluid connection with the peripheral vein and configured to generate an electronic signal associated with the PVP while the circulatory system of the patient is connected to the pump; and

an evaluation unit comprising a computer processor communicatively connected to the PVP sensor to receive the electronic signal and a memory storing non-transitory computer readable instructions that, when executed by the computer processor, cause the evaluation unit to:

obtaining a time-domain PVP signal comprising values of an electronic signal associated with the PVP from the transducer based on a physical phenomenon associated with the PVP of the patient over a sampling period, wherein the sampling period comprises a plurality of time segments including:

(i) one or more active time segments in which the pump is operating; and

(ii) one or more idle time segments in which the pump is not operating;

identifying, based on the evaluation of the values of the time-domain PVP signal, a first plurality of values of the time-domain PVP signal associated with the one or more idle time segments and a second plurality of values of the time-domain PVP signal associated with the one or more active time segments;

generating a filtered time-domain PVP signal based on the first plurality of values and excluding the second plurality of values;

applying a transform to the filtered time-domain PVP signal to generate a frequency-domain PVP signal; and

determining a patient state metric for the patient based on the frequency-domain PVP signal via an equation that takes into account a plurality of harmonic frequencies.

2. The system of claim 1, wherein the pump is a peristaltic IV pump.

3. The system of claim 1, wherein the pump is configured to operate periodically such that the one or more active time segments and the one or more idle time segments periodically alternate.

4. The system of claim 1, wherein the IV tube is disposed between the patient and the pump such that a portion of the pump is in fluid connection with the peripheral vein of the circulatory system of the patient via the IV tube.

5. The system of claim 4, wherein:

the transducer comprises a pressure sensor arranged in fluid connection with the interior of the IV tube; and

the physical phenomenon associated with the PVP is the pressure within the interior of the IV tube.

6. The system of claim 4, wherein the instructions further cause the evaluation unit to:

determining whether the patient state metric indicates that the condition of the patient is abnormal; and

adjusting operation of the pump by changing a rate of flow of fluid from the pump into the circulatory system of the patient when the patient state metric indicates that the condition of the patient is abnormal.

7. The system of claim 1, wherein executable instructions to cause the evaluation unit to generate the filtered time-domain PVP signal comprise:

instructions for causing the evaluation unit to remove the one or more active time segments from the time-domain PVP signal.

8. The system of claim 7, wherein the executable instructions further cause the evaluation unit to generate the filtered time-domain PVP signal by, for each of one or more pairs of the active time segments:

identifying one or more corresponding values within both active time segments in the pair; and

combining the active time segments in the pair by aligning the one or more corresponding values within both active time segments in the pair.

9. The system of claim 1, wherein executable instructions to cause the evaluation unit to generate the filtered time-domain PVP signal comprise instructions to cause the evaluation unit to:

estimating a third plurality of values as replacement values for the one or more active time segments, wherein the third plurality of values are estimated based on the first plurality of values without reference to the second plurality of values; and

generating the filtered time-domain PVP signal by combining the first plurality of values for the idle time segment and the third plurality of values for the active time segment.

10. The system of claim 9, wherein the third plurality of values are estimated by performing at least one of regression analysis, forward-backward slope calculation, bilateral slope detection, and image matched filtering on at least the first plurality of values.

11. The system of claim 1, wherein executable instructions for causing the evaluation unit to determine the patient state metric comprise instructions for causing the evaluation unit to:

identifying a plurality of frequencies associated with local maxima of the frequency-domain PVP signal; and

determining the patient state metric based at least in part on at least one of the plurality of frequencies associated with the local maxima.

12. The system of claim 1, wherein the patient state metric is a blood volume metric indicative of one or more of:

when the blood volume is low, the blood pressure is low,

high blood volume, or

Normal blood volume.

13. An apparatus for monitoring a patient, comprising:

a Peripheral Venous Pressure (PVP) sensor comprising a transducer configured to monitor a physical phenomenon associated with a PVP within a peripheral vein of a circulatory system of the patient while the circulatory system of the patient is connected to a pump; and

an evaluation unit comprising a computer processor communicatively connected to the PVP sensor and a memory storing non-transitory executable instructions that, when executed by the computer processor, cause the evaluation unit to:

obtaining a time-domain PVP signal comprising values of an electronic signal associated with the PVP received from the transducer of the PVP sensor over a sampling period, wherein the sampling period comprises a plurality of time segments including:

(i) one or more active time segments in which the pump is operating, and

(ii) one or more idle time segments in which the pump is not operating;

14. The apparatus of claim 13, wherein:

the time-domain PVP signal comprises a first time series of discrete values;

the filtered time-domain PVP signal comprises a second time series of discrete values; and

the second time series contains a plurality of values of an order of at least one segment within the second time series that corresponds to a plurality of corresponding values of an order of a corresponding segment within the first time series.

15. The device of claim 13, wherein executable instructions to cause the evaluation unit to generate the filtered time-domain PVP signal comprise:

16. The device of claim 13, wherein executable instructions to cause the evaluation unit to generate the filtered time-domain PVP signal comprise instructions to cause the evaluation unit to:

17. A method of monitoring a patient while the patient's circulatory system is connected to a pump using measurements associated with Peripheral Venous Pressure (PVP) within peripheral veins of the patient's circulatory system, comprising:

monitoring, by a transducer, a physical phenomenon associated with the patient's PVP over a sampling period, wherein the sampling period comprises a plurality of time segments including:

(i) one or more active time segments in which the pump is operating, and

(ii) one or more idle time segments in which the pump is not operating;

obtaining, by a processor of an evaluation unit, a time-domain PVP signal comprising values of an electronic signal associated with the PVP from the transducer, based on a physical phenomenon monitored over the sampling period;

identifying, by the processor of the evaluation unit, a first plurality of values of the time-domain PVP signal associated with the one or more idle time segments and a second plurality of values of the time-domain PVP signal associated with the one or more active time segments based on the evaluation of the values of the time-domain PVP signal;

generating, by the processor of the evaluation unit, a filtered time-domain PVP signal based on the first plurality of values and excluding the second plurality of values;

applying, by the processor of the evaluation unit, a transform to the filtered time-domain PVP signal to generate a frequency-domain PVP signal; and

determining, by the processor of the evaluation unit, a patient state metric for the patient based on the frequency-domain PVP signal via an equation that takes into account a plurality of harmonic frequencies.

18. The method of claim 17, wherein generating the filtered time-domain PVP signal comprises:

removing the one or more active time segments from the time-domain PVP signal.

19. The method of claim 17, wherein generating the filtered time-domain PVP signal comprises:

20. The method of claim 17, wherein the third plurality of values are estimated by performing at least one of regression analysis, forward-backward slope calculation, bilateral slope detection, and image matched filtering on at least the first plurality of values.

Claims

(i) one or more active time segments in which the pump is operating; and

(ii) one or more idle time segments in which the pump is not operating;

determining a patient state metric for the patient based on the frequency-domain PVP signal.

2. The system of claim 1, wherein the pump is a peristaltic IV pump.

5. The system of claim 4, wherein:

when the blood volume is low, the blood pressure is low,

high blood volume, or

Normal blood volume.

13. An apparatus for monitoring a patient, comprising:

(i) one or more active time segments in which the pump is operating, and

(ii) one or more idle time segments in which the pump is not operating;

14. The apparatus of claim 13, wherein:

the time-domain PVP signal comprises a first time series of discrete values;

(i) one or more active time segments in which the pump is operating, and

(ii) one or more idle time segments in which the pump is not operating;

determining, by the processor of the evaluation unit, a patient state metric for the patient based on the frequency-domain PVP signal.

removing the one or more active time segments from the time-domain PVP signal.